ICVL – 2008

Virtual Learning – Virtual Reality

Proceedings of the 3
rd

INTERNATIONAL CONFERENCE
ON VIRTUAL LEARNING
October 31 – November 2 , 2008, Constanta, ROMANIA


EDITORS: Marin VLADA, Grigore ALBEANU, Dorin Mircea POPOVICI


Bucharest Uni versi t y Press

FP7 – INFORMATION AND COMMUNICATION TECHNOLOGIES
2010 – Towards a Learning and Knowledge Society – 2030


The ICVL 2008 is held under the auspices of the INTUITION
Consortium-The Network of Excellence in Europe and
National Authority for Scientific Research




UNIVERSITY OF BUCHAREST

www.unibuc.ro



National Authority for Scientific Research – www.mct.ro



OVIDIUS UNIVERSITY OF CONSTANTA

www.univ-ovidius.ro

FACULTY OF MATHEMATICS
AND COMPUTER SCIENCE

www.univ-ovidius.ro/math




The 3
rd
International Conference on Virtual Learning
VIRTUAL LEARNING – VIRTUAL REALITY


VIRTUAL ENVIRONMENTS FOR EDUCATION AND RESEARCH

MODELS & METHODOLOGIES, TECHNOLOGIES, SOFTWARE SOLUTIONS

www.icvl.eu/2008 www.cniv.ro/2008

ICVL 2008 Awards - Sponsored by Intel Corporation
1. Excellence Award "Intel®Education" - USD 500
2. Special Award "Intel®Education" - USD 500

The ICVL Award is offered in recognition of ICVL papers published within in "Proceedings of the
International Conference on Virtual Learning"















ICVL and CNIV Coordinator: Dr. MARIN VLADA


The printing of Proceedings was sponsored by the Ministry of Education and Research,
National Authority for Scientific Research, ROMANIA

Proceedings of the 3
rd

International Conference
On Virtual Learning




October 31- November 2, 2008


MODELS & METHODOLOGIES, TECHNOLOGIES, SOFTWARE SOLUTIONS
















, 2008

ICVL and CNIV Partners:
Dr. Grigore Albeanu, Dr. Mircea Popovici, Prof. Radu Jugureanu
www.icvl.eu www.cniv.ro









© Bucharest University Press
Şos. Panduri nr. 90-92, BUCUREŞTI- 050663; Tel.Fax: 410.23.84
E-mail: editura@unibuc.ro
Web: www.editura.unibuc.ro






Tehnoredactare: Emeline-Daniela AVRAM













ISSN: 1844-8933









M MO OT TT TO OS S



„ „The informatics/computer science re-establishes not only the
unity between the pure and the applied mathematical sciences, the
concrete technique and the concrete mathematics, but also that
between the natural sciences, the human being and the society. It
restores the concepts of the abstract and the formal and makes
peace between arts and science not only in the scientist' conscience,
but in their philosophy as well. .” ”



G Gr r. . C C. . M Mo oi is si il l ( (1 19 90 06 6- -1 19 97 73 3) )
Professor at the Faculty of Mathematics, University of Bucharest,
Member of the Romanian Academy,
Computer Pioneer Award of IEEE, 1996
http://fmi.unibuc.ro/icvl/2006/grcmoisil



”Learning is evolution of knowledge over time”


Roger E. Bohn
Professor of Management and expert on technology management,
University of California, San Diego, USA,
Graduate School of International Relations and Pacific Studies
http://irps.ucsd.edu/faculty/faculty-directory/roger-e-bohn.htm


Welcome to ICVL 2008!



The 3
rd
edition of the International Conference on Virtual
Learning continues bringing together scientists, teachers, students,
managers and psychologists in order to present contributions or to find
out the state of the art in the field of Virtual Learning. The logo for this
edition is "VIRTUAL ENVIRONMENTS FOR EDUCATION AND RESEARCH",
and shows the increasing interest in methodologies, models, techniques,
software tools, content development and quality evaluation of educational
software based on virtual environments. Based on a two-stage selection
of the papers, finally only 46 contributions were selected for oral
presentation and publishing in the ICVL proceedings from 74 proposal
received initially. This reveals the large effort of the Scientific committee
in the double-blind reviewing process in order to keep only relevant, good
and very good contributions.
During ICVL event, an Workshop on EMULACTION project, coordinated
by Dr. Jean-Pierre Gerval, ISEN-Brest (école d'ingénieurs généralistes des
hautes technologies, L'Institut Supérieur de l'Electronique et du Numérique),
France will present the stage of development of a Web Based Environment
in order to enable distributed and co-operative practical activities: groups
of students from different schools and different countries working
together on the same activities. The partners of this important project are
Ovidius University of Constanta, Moncton University in Canada,
Viettronics Technology College in Vietnam, Libanese University at Tripoly
- Liban and Technical University of Moldova at Chisinau, and they are
working together to implement the concept of Distributed Virtual Room
for the EMULACTION web based environment.
As usually, the International Conference on Virtual Learning
(http://www.icvl.eu) is part of an important project sponsored by
Romanian Ministry of Education and Research and SIVECO SA Romania.
For the second year, the organisers invite you at University of Constanta
to attend to all the Conference events: presentations, exhibition, welcome
party, and, also to meet people all around the world, including a lot of
young people from Romania participating at The Seven Edition of the
National Conference on Virtual Learning (http://www.cniv.ro) a jointly
event in these days.


8
The organisers greatly appreciate the interest and the contribution
of INTEL corporation which offers two awards for papers published in the
ICVL proceedings which conforms with ICVL objectives and promotes new
methodologies and information technologies in education. In respects with
ICVL objectives to promote valuable research we mention the publication in
extended version of some of the best papers having the focus on the
convergence of the 3 "C" (Computing, Communication, Control) in the
International Journal of Computers, Communications & Control, an ISI
journal (http://journal.univagora.ro/).

The We hope you will find valuable contributions to the field of virtual
learning and you will enjoy the city of Publius Ovidius Naso (43 BC – 17 AD),
the most widely read and imitated of Latin poets, known to the English-
speaking world as Ovid, which in ten years of banishment by Augustus,
he wrote five books of the Tristia, four of the Epistulae ex Ponto, and the
long curse-poem Ibis.

Welcome to Constanta, Romania!

Dr. Marin Vlada and Dr. Grigore Albeanu,
ICVL and CNIV Projects





GENERAL CONTENTS




About ICVL 2008 ................................................................ 11

About Intel®Education ....................................................... 29

About EMULACTION project .............................................. 31

Invited papers,
Projects – Virtual Environments
for Education and Research ................................................. 33

Section M&M
MODELS & METHODOLOGIES ................................................ 99

Contents of Section M&M .................................................... 281

Sections TECH and SOFT
TECHNOLOGIES and
SOFTWARE SOLUTIONS ......................................................... 287

Contents of Sections TECH and SOFT ................................. 397

Section Intel® Education
LEARNING, TECHNOLOGY, SCIENCE ....................................... 401

Contents of Sections Intel® Education ................................. 437

News and Events
ICVL 2008 Web site ............................................................. 439


The 3
rd
International Conference on Virtual Learning, ICVL 2008

11
About ICVL 2008

ICVL Project – www.icvl.eu

2010 – TOWARDS A LEARNING AND KNOWLEDGE SOCIETY – 2030
VIRTUAL ENVIRONMENTS FOR EDUCATION AND RESEARCH




© Project Coordinator: Ph.D. Marin Vlada, University of Bucharest, Romania
Partners: Ph.D. Prof. Grigore Albeanu, Ph.D. Mircea Dorin Popovici,
Prof. Radu Jugureanu
Sponsors: The Romanian Ministry of Education and Research,
SIVECO Romania, Intel Corporation

ICVL is held under the auspices of:

– The European INTUITION Consortium
– The Romanian Ministry of Education and Research
– The National Authority for Scientific Research
University of Bucharest and Ovidius University of Constanta

12

Conference Organisation

• General Chair Dr. Marin Vlada, Professor of Computer
Science, University of Bucharest, Research Center for
Computer Science (Romania), European INTUITION
Consortium member

• Technical Programme Chair Dr. Grigore Albeanu,
Professor of Computer Science, Spiru Haret University,
Research Center for Mathematics and Informatics
(Romania)


• Associate General Chair Dr. Dorin Mircea Popovici,
Professor of Computer Science, Ovidius University of
Constanta (Romania), CERV- European Center for
Virtual Reality (France)


• Associate General Chair Prof. Radu Jugureanu, AeL
eContent Department Manager, SIVECO Romania SA,
Bucharest, Romania





Scientific Committee/Technical Programme Committee / Executive reviewers
The 3
rd
International Conference on Virtual Learning, ICVL 2008

13
Dr. Grigore
Albeanu
Professor of Computer Science, Spiru Haret University, Research
Center for Mathematics and Informatics, Romania
Dr. Adrian
Adascalitei
Professor of Electrical Engineering Fundamentals, Technical University
"Gh. Asachi", Faculty of Electrical Engineering, Iasi, Romania
Dr. Angelos
Amditis
Research Associate Professor (INTUITION Coordinator,
http://www.intuition-eunetwork.net/), Institute of Communication and
Computer Systems, ICCS- NTUA Microwaves and Optics Lab,
ATHENS, GREECE
Dr. Grigore
Burdea
Professor of Applied Science (Robotics), Rutgers – The State
University of New Jersey, Director, Human-Machine Interface
Laboratory, CAIP Center, USA
Dr. Pierre
Chevaillier
LISYC – Laboratoire d'Informatique des Systèmes Complexes,
CERV – Centre Européen de Réalité Virtuelle (European Center for
Virtual Reality), France, European INTUITION Consortium member
Dr. Mirabelle
D' Cruz
Virtual Reality Applications Research Team (VIRART), School of
Mechanical, Materials and Manufacturing Engineering
(M3),University of Nottingham University, U.K., European
INTUITION Consortium member
Dr. Steve
Cunningham
Noyce Visiting Professor of Computer Science, Grinnell College,
Grinnell, Iowa 50112, USA Department of Computer Science
Dr. Ioan Dzitac
Professor of Computer Science, Executive Editor of IJCCC, Agora
University,Oradea, Romania
Dr. Victor Felea
Professor of Computer Science, “Al.I. Cuza” University of Iasi,
Faculty of Computer Science, Romania
Dr. Horia
Georgescu
Professor of Computer Science University of Bucharest, Faculty of
Mathematics and Computer Science, Romania
Dr. Radu
Gramatovici
Professor of Computer Science University of Bucharest, Faculty of
Mathematics and Computer Science, Romania
Dr. Felix
Hamza-Lup
Professor of Computer Science at Armstrong Atlantic State University,
USA
Dr. Angela Ionita
Romanian Academy, Institute for Artificial Intelligence (RACAI),
Deputy Director, Romania
Olimpius Istrate
Intel Education Manager, Bucharest, Romania
www.intel.com/education
Prof. Radu
Jugureanu
AeL eContent Department Manager, SIVECO Romania SA, Bucharest,
Romania www.siveco.ro
Dr. Bogdan
Logofatu
Professor at University of Buchares, CREDIS Department Manager,
Bucharest, Romania www.unibuc.ro
Dr. Jean-Pierre
Gerval
ISEN Brest (école d'ingénieurs généralistes des hautes technologies),
France, European INTUITION Consortium member
University of Bucharest and Ovidius University of Constanta

14
Dr. Daniel
Mellet-d'Huart
AFPA Direction de l'Ingénierie Unité Veille sur la Réalité Virtuelle
MONTREUIL, European INTUITION Consortium member
Dr. Mihaela Oprea Professor in the Department of Informatics, University of Ploiesti, Romania
Thomas Osburg Intel Education Manager, Europe www.intel.com/education
Dr. Harshada(Ash)
Patel
Virtual Reality Applications Research Team (VIRART)/Human Factors
Group Innovative Technology Research Centre, School of Mechanical,
Materials and Manufacturing Engineering, University of Nottingham,
University Park, Nottingham, U.K., European INTUITION Consortium member
Dr. Dana Petcu
Professor at Computer Science Department of Western University of
Timisoara, Director at Institute e-Austria Timisoara, Romania
Dr. Dorin Mircea
Popovici
Professor of Computer Science, Ovidius University of Constanta,
Romania / CERV– European Center for Virtual Reality (France,
European INTUITION Consortium member)
Dr. Ion Roceanu
Professor of Computer Science, Director of the Advanced Distributed
Learning Department, "Carol I" National Defence University,
Bucharest, Romania
Dr. Maria
Roussou
Virtual Environments and Computer Graphics Lab., Department of
Computer Science, University College London, U.K., European
INTUITION Consortium member
Dr. Ronan
Querrec
CERV – Centre Européen de Réalité Virtuelle (European Center for Virtual
Reality), Laboratoire d'Informatique des Systèmes Complexes, France
Dr. Luca-Dan
Serbanati
Professor of Computer Science, University "Politehnica" of Bucharest,
Romania and Professor at the "La Sapienza" University, Italy,
European INTUITION Consortium member
Dr. Doru Talaba
Professor, “Transilvania” University of Brasov, Product Design and Robotics
Department, Romania, European INTUITION Consortium member
Dr. Leon
Tambulea
Professor of Computer Science, "Babes-Bolyai" University, Cluj-Napoca,
Romania
Dr. Jacques
Tisseau
CERV – Centre Européen de Réalité Virtuelle (European Center for
Virtual Reality), LISYC – Laboratoire d'Informatique des Systèmes
Complexes, France, European INTUITION Consortium member
Dr. Alexandru
Tugui
Professor at “Al. I. Cuza” University of Iasi, FEAA, “Al. I. Cuza”
University Iasi, Romania
Dr. Marin Vlada
Professor of Computer Science, University of Bucharest, Faculty of
Mathematics and Computer Science, Romania, European INTUITION
Consortium member


The 3
rd
International Conference on Virtual Learning, ICVL 2008

15

ICVL 2008 INVITATION

2010 – Towards a Learning and Knowledge Society – 2030


ICVL 2008 – The 3rd International Conference on Virtual Learning
NEWS TECHNOLOGIES IN EDUCATION AND RESEARCH



2010 – Towards a Learning and Knowledge Society – 2030

Virtual Environments for Education and Training, Software and Management for Education

October 31-November 2, 2008 Constanta, ROMANIA
Host: University OVIDIUS Constanta, Faculty of Mathematics and Computer
Science, ROMANIA

Organizers: University of Bucharest and University OVIDIUS Constanta in
cooperation with SIVECO SA company, Bucharest, Romania
Sponsors: National Authority for Scientific Research, SIVECO SA company,
Bucharest, Romania, Intel Coporation
Homepage: http://www.icvl.eu/2008 Email: icvl[at]fmi.unibuc.ro
Deadline for abstracts: June 30, 2008

Description:
At the Lisbon European Council in March 2000, Heads of State and Government
set an ambitious target for Europe to become "the most competitive and dynamic
knowledge-based economy in the world" by 2010. They also placed education
firmly at the top of the political agenda, calling for education and training systems
to be adapted to meet this challenge.

University of Bucharest and Ovidius University of Constanta

16
POST-CONFERENCE: The Organisation Committee will elaborate until the ICVL opening,
the volume with the conference's papers and the CD (with ISBN). Extended versions of
selected papers presented at ICVL will be offered for publishing in the International
Journal of Computers, Communications & Control – http://www.journal.univagora.ro/

AIMS:
– The implementation of the Information Society Technologies (IST) according to
the European Union Framework-Programme (FP7)
– The development of Research, projects, and software for E-Learning, Software
and Educational Management fields;
– To promote and develop scientific research for e-Learning, Educational
Software and Virtual Reality;

SECTIONS:
MODELS & METHODOLOGIES (M&M); TECHNOLOGIES (TECH); SOFTWARE
SOLUTIONS (SOFT)
"Intel® Education" – Learning, Technology, Science (IntelEdu):

Research papers - Major Topics
The papers describing advances in the theory and practice of Virtual Environments
for Education and Training (VEE&T), Virtual Reality (VR), Information and Knowledge
Processing (I&KP), as well as practical results and original applications.
The education category includes both the use of Web Technologies, Computer
Graphics and Virtual Reality Applications, New tools, methods, pedagogy and
psychology, Case studies of Web Technologies and Streaming Multimedia
Applications in Education, experience in preparation of courseware.
Thank you very much for your attention and, please, circulate this call for papers.

Thank you and best regards,
Mail Address:
Str. Academiei nr.14, sector 1, C.P. 010014, Bucuresti, Romania
Tel: (4-021) 314 3508, Fax: (4-021) 315 6990,

Submitted by: Dr. Marin Vlada
Date received: February 08, 2008

Participate
The Conference is structured such that it will:
• provide a vision of European e-Learning and e-Training policies;
• take stock of the situation existing today;
• work towards developing a forward looking approach.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

17
The Conference will consider the perspectives and vision of the i-2010 programme and
how this will stimulate the promotion, and development of e-Learning content, products
and services and the contribution of these to lifelong learning.
Participation is invited from researches, teachers, trainers, educational authorities,
learners, practitioners, employers, trade unions, and private sector actors and IT industry.
Research papers – Major Topics

The papers describing advances in the theory and practice of Virtual Environments for
Education and Training (VEL&T), Virtual Reality (VR), Information and Knowledge
Processing (I&KP), as well as practical results and original applications. The education
category includes both the use of Web Technologies, Computer Graphics and Virtual
Reality Applications, New tools, methods, pedagogy and psychology, Case studies of
Web Technologies and Streaming Multimedia Applications in Education, experience in
preparation of courseware.
Thematic Areas / Sections
• MODELS & METHODOLOGIES (M&M)
• TECHNOLOGIES (TECH)
• SOFTWARE SOLUTIONS (SOFT)
• "Intel® Education" – Learning, Technology, Science (IntelEdu)
General Chair Dr. Marin Vlada, Professor of Computer Science, University of
Bucharest (Romania) / Technical Programme Chair Dr. Grigore Albeanu, Professor of
Computer Science, Spiru Haret University, Research Center for Mathematics and
Informatics (Romania) /
Associate General Chair Dr. Dorin Mircea Popovici, Professor of Computer Science,
Ovidius University of Constanta (Romania)


ICVL 2008 – Announcements and call for papers

• www.intuition-eunetwork.org/ – INTUITION Forum: Conferences, Workshops,
Call for Papers
• www.allconferences.com (E-Learning , Higher Education)
• www.conferencealerts.com – Academic Conferences Worldwide
• http://atlas-conferences.com/ – Database of academic conference announcements
• http://www.xplora.org – The European gateway to science education
• www.papersinvited.com – Powered by CSA / (CSA is a worldwide information
company)
• www.cncsis.ro, www.edu.ro, www.agora.ro, www.ad-astra.ro – romanian sites
University of Bucharest and Ovidius University of Constanta

18
Objectives
2010 – Towards a Learning and Knowledge Society – 2030
At the Lisbon European Council in March 2000, Heads of State and Government set an
ambitious target for Europe to become "the most competitive and dynamic knowledge-
based economy in the world" by 2010. They also placed education firmly at the top of the
political agenda, calling for education and training systems to be adapted to meet this challenge.
Relevant topics include but are not restricted to:
• National Policies and Strategies on Virtual Learning
• National Projects on Virtual Universities
• International Projects and International Collaboration on Web-based Education
• Dot-com Educational Institutions and their Impact on Traditional Universities
• Educational Portals for education and training
• Reusable Learning Objects for e-Learning and e-Training
• Testing and Assessment Issues of Web-based Education
• Academia/Industry Collaboration on Web-based Training
• Faculty Development on Web-based Education
• Funding Opportunities for Projects in Web-based Education
Learning and the use of Information and Communication Technologies (I&CT) will
be examined from a number of complementary perspectives:
• Education – supporting the development of key life skills and competences
• Research – emerging technologies and new paradigms for learning
• Social – improving social inclusion and addressing special learning needs
• Enterprise – for growth, employment and meeting the needs of industry
• Employment – lifelong learning and improving the quality of jobs
• Policy – the link between e-Learning and European / National policy imperatives
• Institutional – the reform of Europe’s education and training systems and how
I&CT can act as catalyst for change
• Industry – the changing nature of the market for learning services and the new
forms of partnership that are emerging

General Objectives
The implementation of the Information Society Technologies (IST) according to
the European Union Framework-Programme (FP6, FP7)
• The implementation of the Bologna Conference (1999) directives for the Romanian
educational system.
• The development of a Romanian Framework supporting the professional and
management initiatives of the educational community.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

19
• The organization of the activities concerning the cooperation between the educational
system and the economical companies to find out an adequate distribution of the
human resources over the job market.
• To promote and implement the modern ideas for both the initial and continuing
education, to promote the team based working, to attract and integrate the young
graduates in the Research and Development projects, to promote and implement
IT&C for initial and adult education activities.

Particular objectives
The development of Research, projects, and software for E-Learning, Software
and Educational Management fields
• To promote and develop scientific research for e-Learning, Educational Software
and Virtual Reality
• To create a framework for a large scale introduction of the e-Learning approaches
in teaching activity.
• To assist the teaching staff and IT&C professionals in the usage of the modern
technologies for teaching both in the initial and adult education.
• To improve the cooperation among students, teachers, pedagogues, psychologists
and IT professionals in specification, design, coding, and testing of the educational
software.
• To increase the teachers' role and responsibility to design, develop and use of the
traditional technologies and IT&C approaches in a complementary fashion, both
for initial and adult education.
• To promote and develop information technologies for the teaching, management
and training activities.
• To promote and use Educational Software Packages for the initial and adult education.


Thematic Areas/Sections

Models & Methodologies (M&M):
• Innovative Teaching and Learning Technologies
• Web-based Methods and Tools in Traditional, Online Education and Training
• Collaborative E-Learning, E-Pedagogy,
• Design and Development of Online Courseware
• Information and Knowledge Processing
• Knowledge Representation and Ontologism
• Cognitive Modelling and Intelligent systems
• Algorithms and Programming for Modelling

Technologies (TECH):
• Innovative Web-based Teaching and Learning Technologies
University of Bucharest and Ovidius University of Constanta

20
• Advanced Distributed Learning (ADL) technologies
• Web, Virtual Reality/AR and mixed technologies
• Web-based Education (WBE), Web-based Training (WBT)
• New technologies for e-Learning, e-Training and e-Skills
• Educational Technology, Web-Lecturing Technology
• Mobile E-Learning, Communication Technology Applications
• Computer Graphics and Computational Geometry
• Intelligent Virtual Environment

Software Solutions (SOFT):
• New software environments for education & training
• Software and management for education
• Virtual Reality Applications in Web-based Education
• Computer Graphics, Web, VR/AR and mixed-based applications for education &
training, business, medicine, industry and other sciences
• Multi-agent Technology Applications in WBE and WBT
• Streaming Multimedia Applications in Learning
• Scientific Web-based Laboratories and Virtual Labs
• Software Computing in Virtual Reality and Artificial Intelligence
• Avatars and Intelligent Agents

Research papers – Major Topics
The papers describing advances in the theory and practice of Virtual Environments
for Education and Training (VEL&T), Virtual Reality (VR), Information and Knowledge
Processing (I&KP), as well as practical results and original applications. The
education category includes both the use of Web Technologies, Computer
Graphics and Virtual Reality Applications, New tools, methods, pedagogy and
psychology, Case studies of Web Technologies and Streaming Multimedia
Applications in Education, experience in preparation of courseware.

Topics of interest include but are not limited to:
Virtual Environments for Learning (VEL):
• New technologies for e-Learning, e-Training and e-Skills
• New software environments for education & training
• Web & Virtual Reality technologies
• Educational Technology and Web-Lecturing Technology
• Advanced Distributed Learning (ADL) technologies
• Innovative Web-based Teaching and Learning Technologies
• Software and Management for Education
• Intelligent Virtual Environment

Virtual Reality (VR):
• Computer Graphics and Computational Geometry
• Algorithms and Programming for Modeling
• Web & Virtual Reality-based applications
The 3
rd
International Conference on Virtual Learning, ICVL 2008

21
• Graphics applications for education & training, business, medicine, industry
and other sciences
• Scientific Web-based Laboratories and Virtual Labs
• Software Computing in Virtual Reality

Knowledge Processing (KP):
• Information and Knowledge Processing
• Knowledge Representation and Ontologism
• Multi-agent Technology Applications in WBE and WBT
• Streaming Multimedia Applications in Learning
• Mobile E-Learning, Communication Technology Applications
• Cognitive Modelling, Intelligent systems
• New Software Technologies, Avatars and Intelligent Agents
• Software Computing in Artificial Intelligence


Fundamentals | Educational Technology

Deploying Education Environments for the 21st Century (Robert Fogel and Steve Gish, Intel
Corporation) (.pps) – http://www.intel.com/education
Educational Technology that Talks – http://www.edtechtalk.com
The Best Virtual Reality Information on Internet – http://vresources.org/
Kaleidoscope – the European research network shaping the scientific evolution of technology
enhanced learning – www.noe-kaleidoscope.org/pub/

• Project Zero – Project Zero is an educational research group at the Graduate School of
Education at Harvard University | www.pz.harvard.edu
• Project Zero eBookstore - Featured Publications from Project Zero |
www.pz.harvard.edu/ebookstore
For Gardner, intelligence is:
– the ability to create an effective product or offer a service that is valued in a culture;
– a set of skills that make it possible for a person to solve problems in life;
– the potential for finding or creating solutions for problems, which involves gathering new knowledge.
Five Minds for the Future" (NEW BOOK) Harvard Business School Press
Gardner's newest book, Five Minds for the Future outlines the specific cognitive abilities that will
be sought and cultivated by leaders in the years ahead.
They include:
1. The Disciplinary Mind: the mastery of major schools of thought, including science, mathematics,
and history, and of at least one professional craft.
2. The Synthesizing Mind: the ability to integrate ideas from different disciplines or spheres into a
coherent whole and tocommunicate that integration to others.
3. The Creating Mind: the capacity to uncover and clarify new problems, questions and phenomena.
4. The Respectful Mind: awareness of and appreciation for differences among human beings and
human groups.
5. The Ethical Mind: fulfillment of one's responsibilities as a worker and as a citizen.
University of Bucharest and Ovidius University of Constanta

22
• History of Virtual Learning Environments – "Integrated Learning Systems" (ILS), "Computer
Assisted Instruction" (CAI), "Computer Based Training" (CBT),"Computer Managed
Instruction" (CMI), "Interactive Multimedia Instruction" (IMI), "Technology Enhanced
Learning" (TEL), "Technology Based Learning" (TBL), and "Web Based Training" (WBT)
(Reference: http://en.wikipedia.org/)
• Information Society Technologies – The four waves of information technologies
(Reference: Vlada, M., Tugui, Al., The First International Conference on Virtual
Learning – ICVL 2006, october 27-29, pp. 69-82, Proceedings of ICVL 2006 and CNIV 2006, 2006.)
• The terminology used in the fields of Virtual Learning
(Reference: Anohina A., Analysis of the terminology used in the field of virtual learning,
Educational Technology & Society, 8 (3), 91-102, http://www.ifets.info/journals/8_3/9.pdf, 2005.)
• The Evolution of Technological Knowledge
(Bohn, Roger E. (2005). "From Art to Science in Manufacturing: The Evolution of
Technological Knowledge." Foundations and Trends in Technology, Information and
Operations Management 1(2): 1-82.)
• Advanced Distributed Learning – ADL – Creating the knowledge environment of the
future – www.adlnet.gov (This is an official Web site of the U.S. Government)
• ADL Technologies: Sharable Content Object Reference Model (SCORM); Content Object
Repository Discovery and Registration Architecture (CORDRA); Simulations; Intelligent Tutoring
• SCORM Technologies – Sharable Content Object Reference Model ("SCORM is a collection
of specifications adapted from multiple sources to provide a comprehensive suite of e-learning
capabilities that enable interoperability, accessibility and reusability of Web-based learning
content" – www.adlnet.gov)
• AeL Educational, AeL Enterprise – Computer-assisted learning system (e-Learning for
schools and universities): Learning Management – AeL LMS (Learning Management
System); eContent Management – AeL LCMS (Learning Content Management System);
Interactive Multimedia Educational Content – AeL eContent, eContent demo. AeL Enterprise:
AeL Enterprise is a modern learning and management instrument dedicated to supporting
personnel training within the company frame: it is devised for the direct Computer Assisted
Learning (CAL), as well as for the remote / non assisted training (Computer Based Training)

Resources

Educational Technology That Talks – http://www.edtechtalk.com
Kaleidoscope – the European research network shaping the scientific evolution of technology enhanced
learning – www.noe-kaleidoscope.org/pub/
The Best Virtual Reality Information on Internet – http://vresources.org/
Career Opportunities in Academe/Research: University500 – http://www.university500.com
Ad Astra – An Online Project for the Romanian Scientific Community – http://www.ad-astra.ro
1. Sixth Framework Programme (FP6) – http://fp6.cordis.lu/fp6/home.cfm
2. Seventh Framework Programme (FP7) – http://www.cordis.lu/fp7/
3. European Research Area (ERA) – http://www.cordis.lu/era/
4. Information Society Technologies (IST) – http://www.cordis.lu/ist/
5. Information and Communication Technologies (ICT) – http://cordis.europa.eu/fp7/ict/
The 3
rd
International Conference on Virtual Learning, ICVL 2008

23
6. EPISTEP – EPISTEP is an innovative project supported by the EU "Research and Innovation"
(FP6,FP7) – www.epistep.org | European Technology Platforms (ETP) – eMobility, ARTEMIS,
ENIAC, NEM | Networked and electronic media platform – http://www.nem-initiative.org/
7. Eupope's Information Society – http://europa.eu.int/information_society/
8. Eupopean Institute for E-Learning (EifEL) – http://www.eife-l.org/
9. eEurope 2005 – http://europa.eu.int/information_society/eeurope/
10. eEurope+ – http://europa.eu.int/information_society/eeurope/plus/
11. i2010 European Information Society in 2010 – http://europa.eu.int/information_society/eeurope/i2010/
12. European e-Skills 2006 Conference – Towards a Long Term e-Skills Strategy: http://eskills.
cedefop.europa.eu/conference2006/index.asp

13. Intuition Project-Network Of Excellence Focused on Virtual Reality and Virtual Environments
Applications for Future Workspaces – http://www.intuition-eunetwork.net/
14. European Mathematical Society (EMS) – http://www.emis.de/
15. Integrating New Technologies intothe Methods of Education – http://www.intime.uni.edu/
16. Xplora – European gateway to science education – http://www.xplora.org/
17. European Schoolnet – http://www.eun.org/
18. Virtual Learning Systems – http://eservices.nysed.gov/vls/
19. Eastern Europe eWork – http://www.e3work.com/
20. VResources – The Best Virtual Reality Information on Internet: Applications; Events; Discussion
forums; Library; VR News | http://vresources.org/

21. Advanced Distributed Learning – ADL – Creating the knowledge environment of the
future – www.adlnet.gov (This is an official Web site of the U.S. Government)
22. ADL Technologies: Sharable Content Object Reference Model (SCORM); Content Object
Repository Discovery and Registration Architecture (CORDRA); Simulations; Intelligent Tutoring
23. SCORM Technologies – Sharable Content Object Reference Model ("SCORM is a collection
of specifications adapted from multiple sources to provide a comprehensive suite of e-learning
capabilities that enable interoperability, accessibility and reusability of Web-based learning
content" – www.adlnet.gov)
24. W3C – The World Wide Web Consortium (W3C) – www.w3.org | Tim Berners-Lee,
inventor of the World Wide Web
25. International World Wide Web Conference Committee (IW3C2) – http://www.iw3c2.org/ |
15th International World Wide Web Conference, Edinburgh Scotland

26. Moodle Services – Moodle is a course management system designed to help educators who want to
create quality online courses; "Moodle is a real gift to forward thinking educators" – www.moodlle.com
27. Drupal – Drupal is a free software package that allows an individual or a community of users to
easily publish, manage and organize a wide variety of content on a website; Drupal.org is the
official website of Drupal, an open source content management platform – www.drupal.org
University of Bucharest and Ovidius University of Constanta

24
28. Scalable Vector Graphics (SVG) – XML Graphics for the Web; SVG is a language for describing
two-dimensional graphics and graphical applications in XML (Mozilla SVG Project) – www.w3.org/
Graphics/SVG/ | www.svg.org | www.adobe.com/svg/ | www.w3schools.com/svg/
29. AJAX – Ajax (also known as AJAX), shorthand for "Asynchronous JavaScript and XML," is a
development technique for creating interactive web applications (AJAX is a type of programming
made popular in 2005 by Google) – http://en.wikipedia.org/wiki/AJAX | http://ajax.asp.net/
30. FLEX – Adobe Flex is a framework that helps you build dynamic, interactive rich Internet applications
(www.flex.org/) – http://en.wikipedia.org/wiki/Adobe_Flex | www.adobe.com/products/flex/
31. KP (KnowledgePresenter) – Create fully interactive SCORM compliant e-learning
lessons – http://knowledgepresenter.com/
32. SOFTAKE – Software, programs, downloads (Windows, Linux, Mac) – http://www.softake.com/

33. THE COMPUTER GRAPHICS SOCIETY ( C G S ) | International Conference on Computer
Animation and Social Agents – CASA 2005 | CASA 2006
34. ACM SIGGRAPH – Computer Graphics and interactive techniques – http://www.siggraph.org/
35. CGAL – Computational Geometry Algorithms Library - http://www.cgal.org
36. The Human Interface Technology Lab (HITLab, University of Washington) – www.hitl.
washington.edu/ | Virtual Environments in Education and Training (Research Projects) – Dr.
William D. Winn (What We Have Learned About VR and Learning and What We Still Need
to Study. In Proceedings of Laval Virtual 2005)
37. Online Educa Berlin – 12th International Conference on Technology Supported Learning &
Training: http://www.online-educa.com/
38. ACM Symposium on Virtual Reality Software and Technology (VRST) – The conference
will take place in Cyprus 1st-3rd of November 2006 (Cyprus2006) | The first VRST was held
in Singapore in 1994 and since then it has been held in Japan, Hong Kong, Switzerland,
Taiwan, England, Korea, Canada and the US.(www.vrst.org/)
39. How People Learn (the National Academy of Sciences, USA) – http://newton.nap.edu/
html/howpeople1/
40. Simulation Interoperability Standards Organization (SISO) – http://www.sisostds.org

41. Romanian Academy, ROINTERA project – http://www.rointera.ro
42. eLearning Conference, Towards a Learning Society – http://www.elearningconference.org
43. e-Learning Centre UK – http://iet.open.ac.uk/research/confdiary/
44. PROLEARN virtual competence centre – http://www.prolearn-online.com/
45. PCF5 – The Fifth Pan-Commonwealth Forum on Open Learning, 13-17 July 2008, University
of London, UK | www.london.ac.uk/pcf5 | www.col.org/
46. WikiEducator – free eLearning content that anyone can edit and use | www.wikieducator.org
47. EdTechTalk – EdTechTalk is a webcasting network of educators dedicated to helping those involved
in educational technology explore, discuss, and collaborate in its use | http://www.edtechtalk.com/
48. Commonwealth of Learning – COL is an intergovernmental organisation created by
Commonwealth Heads of Government to encourage the development and sharing of open
learning and distance education knowledge, resources and technologies. | www.col.org
The 3
rd
International Conference on Virtual Learning, ICVL 2008

25
49. Innovative Educators – Innovative Educators is dedicated to providing superior conferences
and training sessions focused on the most critical and relevant issues facing educators today |
www.innovativeeducators.org/

50. IIIS – the International Institute of Informatics and Systemics – www.iiis.org/iiis/ | Conferences and
Symposia being organized by IIIS | http://www.iiis.org/iiis/IIISConferences.asp
51. IADIS – International Association for Development of the Information Society – http://www.iadis.
org/es2005/
52. ESPIT – eHealth and eInclusion – http://www.epist.org/
53. VRMI – Virtual Reality Medical Institute, Europe, Brussels – http://www.vrphobia.eu/ | Journal
of CyberTherapy and Rehabilitation (JCR), Annual Review of CyberTherapy and Telemedicine:
The International Association of CyberTherapy & Rehabilitation [ Publications ]
54. Conference Mobile Learning 2005 – http://www.iadis.org/ml2005
55. Winter School of Computer Graphics (WSCG) – http://wscg.zcu.cz
56. IEEE, Computer Society – http://www.computer.org/
57. Springer-Verlag-London – http://www.springer.de/, http://www.springerlink.com/
58. Kluwer Academic Publishers – http://www.kluweronline.com/
59. Science Direct/Elsevier B.V. – http://www.sciencedirect.com/
60. Open Access Initiative – Open Access Journals | OAI is a new paradigm in scholarly
publishing. It aims to promote models that ensure free and unrestricted access to scholarly &
research journals | www.openj-gate.com
61. Computer Animation and Virtual Worlds – InterScience, Journal published by JOHN WILEY
62. The Journal of Visualization and Computer Animation – InterScience, Journal published by
JOHN WILEY
63. International Journal of Knowledge and Learning (IJKL) – http://www.inderscience.
com/browse/index.php?journalCODE=ijkl
64. Virtual Retrospect 2005 – http://www.virtualretrospect.estia.fr/index.htm
65. IARIA – International Academy, Research, and Industry Association (Silicon Valley,
USA) – www.iaria.org/
66. IATED – The International Association for Technology, Education and Development – http://www.
iated.org/
67. IJ-SoTL – New International Journal for the Scholarship of Teaching and Learning (Georgia
Southern University, Georgia, USA)[read more]
68. ICONS 2007 – 1st International Conference on Network Security and Workshop (Erode
Sengunthar Engineering College, India)[read more]
69. HCI2007 – 12th International Conference on Human-Computer Interaction: http://www.hcii2007.org/
70. CISSE 2006 Online E-Conference – The Second International Joint Conferences on Computer,
Information, and Systems Sciences, and Engineering: http://www.cisse2006online.org
71. INSEAD – The Centre for Advanced Learning Technologies (CALT)-France, The Centre for
Advanced Learning Technologies, is one of the well-established Centres of Excellence at
INSEAD. Research focuses on advanced learning systems | http://www.calt.insead.edu/ |
www.insead.edu
72. Laval Virtual ReVolution 2007 – 9th Virtual Reality International Conference, April 18-22th
2007, Laval, France (www.laval-virtual.org) | Demonstrations | Awards 2007 | Student
competitions | VRIC-Virtual Reality International Conference

University of Bucharest and Ovidius University of Constanta

26
73. ACI – ACADEMIC CONFERENCES INTERNATIONAL (www.academic-conferences.org)
| Conferences | e-Journals | Publications | Training&Seminars
74. ICEL 2007 – The International Conference on e-Learning (ICEL), Columbia University, New
York, USA, 28-29 June 2007
75. ECEL 2007 – The European Conference on e-Learning (ECEL), 4-5 October 2007,
Copenhagen Business School, Copenhagen, Denmark
76. Distance Teaching & Learning – The Annual Conference on Distance Teaching & Learning
(LEARN), August 8-10, 2007, Madison Wisconsin, USA
77. ICTL 2007 – International Conference on Teaching and Learning (ICTL), November 15-16, 2007,
Putrajaya , Malaysia
78. IVA 07 – International Conference on Intelligent Virtual Agents(IVA), 17
th
-19
th
September 2007,
Paris, France
79. Scalable Vector Graphics – International Conference on Scalable Vector Graphics (SVGOPEN),
September 4-7, 2007, Tokyo, Japan
80. mLearn 2007 – International Conference on mobile Learning (mLearn), 16-19 October 2007,
Melbourne, Australia

81. ICE 2007 – International Conference on Education (ICE), 21 may 2007, Uniuversity Brunei
Darussalam, China
82. KES 2007 – Artificial Intelligence Applications in Digital Content (KES), September 12-14 2007,
Vietri sul Mare, Italy
83. EC-TEL 2007 – European Conference on Technology Enhanced Learning (EC-TEL),
17-20 September 2007, Crete, Greece
84. CGV 2007 – IADIS International Conference on Computer Graphics and Visualization
(CGV), 6-8 July, 2007 Lisbon, Portugal
85. CGI 2007 – Computer Graphics International (CGI), May 30th - June 2nd, 2007, Petropolis, Brazil
86. SIGGRAPH 2007 – The 34
th
International Conference on Computer Graphics and Interactive
Techniques (SIGGRAPH), 5-9 August 2007, San Diego, California, USA
87. ACVIT 2007 – International Conference on Advances in Computer Vision and Information
Technology (ACVIT), 28-30 November 2007, Aurangabad, Maharashtra, India
88. DC 2007 – International Conference on Dublin Core and Metadata Applications(DC),
27 to 31 August 2007, Singapore
89. ICWL 2007 – The 6th International Conference on Web-based Learning, 15-17 August 2007,
University of Edinburgh, United Kingdom (www.hkws.org/events/icwl2007/)
90. Informatics Europe – The Research and Education Organization of Computer Science and
IT Departments in Europe (www.informatics-europe.org/)
91. European Computer Science Summit – 3
rd
Annual Informatics Europe Meeting 2007
(http://kbs.cs.tu-berlin.de/ecss/), October 8-9 2007, Berlin

92. KCPR 2007 – The 2nd International Symposium on Knowledge Communication and Peer
Reviewing (http://www.info-cyber.org/kcpr2007/), July 12-15, 2007 – Orlando, Florida, USA
93. CITSA 2007 – The 4th International Conference on Cybernetics and Information Technologies,
Systems and Applications (http://www.info-cyber.org/citsa2007/), July 12-15, 2007 – Orlando,
Florida, USA
The 3
rd
International Conference on Virtual Learning, ICVL 2008

27
94. CCCT 2007 – The 5th International Conference on Computing, Communications and Control
Technologies (http://www.info-cyber.org/ccct2007/), July 12-15, 2007 – Orlando, Florida, USA
95. WCECS 2007 – The World Congress on Engineering and Computer Science 2007 | The WCECS
2007 is composed of the following 15 conferences (San Francisco, USA, 24-26 October, 2007)
96. ICEIT 2007 – The International Conference on Education and Information Technology 2007 |
International Association of Engineers (IAENG) (San Francisco, USA, 24-26 October, 2007)
97. ICIMT 2007 – The International Conference on Internet and Multimedia Technologies 2007
(San Francisco, USA, 24-26 October, 2007)
98. ICMLDA 2007 – The International Conference on Machine Learning and Data Analysis 2007
(San Francisco, USA, 24-26 October, 2007)
99. VRST 2007 – ACM Virtual Reality Software and Technology, Nov 5-7, University of Irvine,
USA | http://www.ics.uci.edu/computerscience/vrst/
100. ICMLA 2007 – The 2007 International Conference on Machine Learning and Applications |
www.icmla-conference.org/icmla07/ (Cincinnati, Ohio USA on Dec 13-15, 2007) |
Association for Machine Learning and Applications (ALMA) | www.cs.csubak.edu/

101. ASTD – American Society for Training & Development (www.astd.org/) | ASTD is the
world’s largest association dedicated to workplace learning and performance professionals |
ASTD 2007, ASTD 2007 International Conference & Exposition – June 3-6, 2007
102. mark steiner – www.marksteinerinc.com/ | mark steiner, inc. is a learning consulting company
specializing in technology-based learning, Chicago, USA
103. LearnLab – The Pittsburgh Science of Learning Center's LearnLab (www.learnlab.org/)
104. i-math – What you need, when you need it (http://www.i-math.com/) | i-Math was incorporated
in 2001 as an organization dedicated to delivering innovative high precision mathematical and
control software solutions to the Educational, R&D, Engineering and Manufacturing
industries in the ASEAN Region | http://www.i-math.com.sg/
105. ICCMSE 2007 – The International Conference of Computational Methods in Sciences and
Engineering 2007, Corfu, Greece , 25-30 September 2007 (http://www.iccmse.org/)

106. The Wolfram Demonstrations Project – The Wolfram Demonstrations Project is an open-code
resource that uses dynamic computation to illuminate concepts in science, technology, mathematics,
art, finance, and a remarkable range of other fields (http://demonstrations.wolfram.com/)
107. Mathematica Technology (Wolfram Research Inc.) – http://www.wolfram.com/
108. MathDL – The Mathematical Sciences Digital Library – http://mathdl.maa.org/
109. MAA – Journal of Online Mathematics and its Applications – http://mathdl.maa.org/
mathDL/4/
110. MAA – Digital Classroom Resources – http://mathdl.maa.org/mathDL/3/
111. Mathematica in Education and Research – http://www.ijournals.net/
112. Maplesoft Canada – http://www.maplesoft.com/
113. IBM-Academic Resource– http://www.alphaworks.ibm.com/academic/
114. Microsoft-Training, eLearning, Career, Events – http://msdn.microsoft.com/tce/
115. Intel-Software Development – http://www.intel.com/
116. Sun Microsytems-Training – http://www.sun.com/training/
117. World Summit on the Information Society – http://www.itu.int/wsis/
University of Bucharest and Ovidius University of Constanta

28

Atlas Conferences – Atlas Conferences has a database of academic conference announcement
www.conferencealerts.com – Academic Conference Worldwide
www.confabb.com – The Conference Community
www.papersinvited.com – Powered by CSA (CSA is a worldwide information company)
AllConferences.Com – Directory of Conferences, conventions, exhibits, seminars, workshops,
events, trade shows and business meetings. Includes calendar, dates, location, web site,
contact and registration information.


ICVL 2008 Web site: http://www.icvl.eu/20008



The 3
rd
International Conference on Virtual Learning, ICVL 2008

29
About Intel®Education

Evolution of Education Environments
Deploying Education Environments for the 21st Century



WEB: www.intel.com/education | www.intel.com/worldahead |
www.classmatepc.com



Deploying Education Environments for the 21st Century (Robert Fogel and Steve Gish,
Intel Corporation) (.pps)

"In today’s economy, the most important resource is no longer labour, capital or land; it
is knowledge.” Peter Drucker


Classroom of Tomorrow
Objectives

Share Intel’s worldwide best practices for education; Education solution towards
21st Century challenges; Identify key ingredients of 1:1 education solution
University of Bucharest and Ovidius University of Constanta

30
• Develop 21st-century skills: media literacy, critical and systems thinking, problem
solving, collaboration, self-direction, global awareness, and civic literacy
• Develop ICT skills: Word processing, online collaboration, Internet research,
multimedia production, etc.
• Improve student access to information: Intranet and Internet connectivity
• Enhance school productivity: Teacher and administrator efficiency
• Improve teaching practice: Improve teachers’ subject knowledge and improve
pedagogical practices, and assist in planning objectives, structuring lessons, etc.
• Improve students’ conceptual understanding: Use dynamic audiovisual representations
to explain concepts and complex information
• Facilitate collaboration: Group projects and improve communication among teachers,
students, parents, and administrators
Education Objectives for the 21st Century

In Terms of the Student:
• Improve the education process
• Improve the education environment
• Prepare students for higher education
• Prepare students thrive in today's global economy
• 21st century skills development
In Terms of a Country or Region:
• Global economic competitiveness
• Grow economy and retain talent pool
• Improve social development
Intel® Education - Learning, Technology, Science

• Digital Curriculum, collaborative rich-media applications, student software,
teacher software
• Improved Learning Methods, interactive and collaborative methods to help
teachers incorporate technology into their lesson plans and enable students to
learn anytime, anywhere
• Professional Development, readily available training to help teachers acquire the
necessary ICT skills
• Connectivity and Technology, group projects and improve communication
among teachers, students, parents and administrators
The 3
rd
International Conference on Virtual Learning, ICVL 2008

31

About EMULACTION Project
Human distributed activities through 3D virtual spaces
ICVL Workshop – EMULACTION Project


https://intranet.iseb.fr/emulaction/

Workshop event run in association with the ICVL 2008
(oct. 31-nov. 2, 2008, Ovidius University of Constanta, Romania).

EMULACTION:
Environnement MULtimodal pour Activités Coopératives Transnationales
de formatION (Multimodal Environment for Transnational and Cooperative
Training Activities)

• This project aims at improving students engineering skills especially when the
actors of the project, the tasks to be achieved and the knowledge are distributed
on several different countries.
• We propose to develop a Web Based Environment in order to enable distributed
and co-operative practical activities: groups of students from different schools
and different countries working together on the same activities. The architecture
University of Bucharest and Ovidius University of Constanta

32
of this Web based Environment will implement the concept of Distributed Virtual
Room. According to the work to be carried out by students, a teacher configures
one (or several) virtual room where a group of students will have to work.

Project COORDINATOR:

• Dr. Jean-Pierre Gerval, ISEN-Brest (école d'ingénieurs généralistes des hautes
technologies, L'Institut Supérieur de l'Electronique et du Numérique), France,
European INTUITION Consortium member, http://www.isen.fr

Partners

1 – OVIDIUS University of Constanta, Constanta, Romania | http://www.univ-ovidius.ro
2 – Moncton University, Canada | http://www.umoncton.ca
3 – Viettronics Technology College, Haiphong – Vietnam | http://www.caodangvtc.edu.vn
4 – Libanese University, Tripoli - Liban | http://www.ul.edu.lb/
5 – Technical University of Moldova, Chisinau, Moldova | http://www.utm.md/en/

Global goal
• The most of the universities develop their international relationships, especially
in order to assure their graduates mobilities. But in a modialisation context,
which is synonim with international transfer of work-resources or knowledge, a
very small part of institutions are able to sensibilise their graduates to the real
chalenges brought by this kind of relationships.
• This why the main goal of EMULACTION is to augment the competence of the
graduates in technique enginee, especially at the project partners, by distributing
the tasks to be completed as well as the knowledge for it.

Specific goal(s)
• In order to reach the main goal the teachers from the partners institution have to
desing an develop some very specific practical works. This suppose the existence
of some specific software tools and a very carreful observation of ergonomics
and easy to use of the resulted training virtual environments.
• In the near future, the EMULACTION project may became a valuable student-
oriented behavioral DB for the future trainers.

Contact persons

• Jean-Pierre Gerval, ISEN-Brest, France
• Dorin-Mircea Popovici, Ovidius University, Constanta, Romania
• Habib Hamam, Moncton University, Canada
• Song Phuong Nguyen, Viettronics Technology College, Haiphong – Vietnam
• Ammar Assoum, Libanese University, Tripoli – Liban
• Valeriu Dulgheru, Technical University of Moldova, Chisinau, Moldova
The 3
rd
International Conference on Virtual Learning, ICVL 2008

33







INVITED PAPERS

Projects


2010 – TOWARDS A KNOWLEDGE SOCIETY – 2030

VIRTUAL ENVIRONMENTS FOR EDUCATION AND RESEARCH

Professional Learning and Knowledge Society
University of Bucharest and Ovidius University of Constanta

34
Number Paper and Authors Page
1.
Virtual Lab: Discovering through Simulation

Jean-Pierre Gerval, Yann Le RU
Institut Supérieur de l’Electronique et du Numérique – Brest
20 rue Cuirassé Bretagne – CS 42807 – 29228 BREST cedex 2 – FRANCE

35
2.
Simulation and Training with Haptic Feedback – A Review

Simona Clapan
1
, Felix G. Hamza-Lup
1

(1) Computer Science, Armstrong Atlantic State University
Savannah, GA 31419, USA
45
3
INTERGEO – Interoperable Interactive Geometry for Europe

Christian Mercat
1
, Paul Libbrecht
2
,
Sophie SouryLavergne
3
, Jana Trgalova
3
(1) I3M, LIRMM, Univ. Montpellier 2, France
(2) German Research Center for Artif. Intel. (DFKI), Saarbrücken, Germany
(3) National Institute for Pedagogical Research (INRP), Lyon, France
53
4
Simulation Models for Virtual Reality Applications

Grigore Albeanu
Spiru Haret University, 13, Ion Ghica Str, RO-30045, ROMANIA
E-mail: galbeanu@gmail.com
63
5
Modeling of Errors Realized by a Human Learner
in Virtual Environment for Training

Thanh Hai Trinh
1, 2
, Cédric Buche
1
,
Jacques Tisseau
1

Université Européenne de Bretagne – ENIB – LISyc – CERV
Technopôle Brest-Iroise, 29238 Brest Cedex 3, FRANCE
(2) Institut de la Francophonie pour l’Informatique
42 Ta Quang Buu, Ha Noi, VIETNAM
E-mail: tthai@ifi.edu.vn; buche@enib.fr; tisseau@enib.fr
71
6
Architecture and Working Principles of the Concept
Map Based Knowledge Assessment System

Marks Vilkelis
1
, Alla Anohina
1
, Romans Lukashenko
1

(1) Department of Systems Theory and Design, Riga Technical University
1, Kalku Str., Riga, LV-1658, LATVIA
E-mail: {markvilkel@inbox.lv, alla.anohina@rtu.lv, LREXPress@inbox.lv}
81
7
Measurement and Control of Statistics Learning Processes
based on Constructivist Feedback and Reproducible
Computing

Patrick Wessa
K.U.Leuven Association, Lessius Dept. of Business Studies, Belgium
E-mail: patrick@wessa.net
91

The 3
rd
International Conference on Virtual Learning, ICVL 2008

35
Virtual Lab: Discovering through Simulation

Jean-Pierre Gerval, Yann Le RU

Institut Supérieur de l’Electronique et du Numérique - Brest
20 rue Cuirassé Bretagne -

CS 42807 - 29228 BREST cedex 2 - FRANCE
Tel: +33 (0)2 98 03 84 07, Fax: +33 (0)2 98 03 84 10
E-mail: {jean-pierre.gerval, yann.le-ru}@isen.fr


Abstract
This paper sets out the design and the implementation of a Virtual Tutor. This
Virtual Tutor is an avatar that “lives” in a distributed virtual reality application
dedicated to practical activities in electronics: circuit design and simulation. The
simulation of the circuit is done using the SPICE programme that is a general-purpose
circuit simulation programme for non-linear dc, non-linear transient, and linear ac
analyses. The implementation is based on VRML (Virtual Reality Modeling Language)
and Java as languages and Cortona VRML plug-in from ParallelGraphics. The
distribution of virtual worlds is obtained using DeepMatrix as environment server.
Teachers use Concept Maps to design the behaviour of the Virtual Tutor. The control of
the avatar is done using JESS (Java Expert System Shell). We describe in this paper
a method that enables the creation of a Knowledge Base from a Concept Map.

Keywords: Distributed Virtual Environments, Virtual Reality Modeling Language,
Java, Concept Maps, Web-based Training.


1. Introduction

Our Virtual Lab. has been experimented with a group of 40 students. This group
represented half a class of 80. The target for this group was to prepare practical activities
using the Virtual Lab. That is to say using virtual components and simulation by means of
the SPICE programme (Gerval and Le RU, 2006).
The other half was preparing practical activities as usual. That is to say using paper
and pens!
All these students were beginners in the field of electronics.
The main functionalities of the Virtual Lab. had been laid out to students during a
short lesson. Then they had to prepare the practical activity by themselves. The students
were expected to study various circuits that implemented operational amplifiers.
Our main goal through this experiment was to assess the relevance of the Virtual
Lab. in the framework of the preparation of practical activities in electronics. Also, as we
had split students into two different groups, we were expecting to make comparisons
about the results of these two groups during practical activities. Benefits from Virtual
Lab. vary according to the students’ behaviour. Students who are eager to work get better
benefits from the Virtual Lab. while the others get only lower gains.
Since the Virtual Lab. resource has been constantly available, well-motivated
students have been encouraged not only to work the courses but also to look further.
University of Bucharest and Ovidius University of Constanta

36
Using the Virtual Lab. is a real added value for these students. As regards the other
students, the Virtual Lab. remains a working tool like others. In this case, the Virtual Lab.
does not increase schoolwork motivation.
On the other hand, this experiment emphasizes the fact that autonomous work using
the Virtual Lab. cannot be applied so simply with the two types of student populations:
naturally autonomous profiles and dependent profiles.
This state of fact is confirmed by their results:
Naturally autonomous profiles are those who succeed better;
Dependent profiles try to get a benefit to escape teacher monitoring.
In order to avoid that students who have a “dependent profile” escape teacher
monitoring, we have decided to implement a Virtual Tutor. The main idea is to give
students the feeling they are working in an autonomous way. But in the fact they are
monitored and this way they can get a feedback about what they are doing.


2. Virtual world description

2.1. Basic components

The implementation of the virtual world is based on VRML (Virtual Reality
Modeling Language) (VRML). Until now, we have implemented six different types of
components (Fig. 1. and Fig. 2.), which are resistors, capacitors, inductors, diodes,
transistors and operational amplifiers.


Figure 1. Passive components Figure 2. Active components

Students can choose a value for resistors or capacitors by selecting the right colours
on the components according to colour codes. Concerning the other components, a menu
has been implemented that enables students to choose a value.


2.2. Designing a circuit

Components are inserted into the virtual world by clicking on the corresponding
icon (Fig. 3.). Students can move (or rotate) components by means of virtual axis (Fig. 4.)
that represent the directions of the movement. After they have put components on the
virtual PCB (Printed Circuit Board), students can build their circuit by clicking on
The 3
rd
International Conference on Virtual Learning, ICVL 2008

37
components’ pins behind the virtual PCB (Fig. 5.). A link is created in the virtual world.
A black line is drawn between components’ pin. This line symbolizes a connection
between two components.


Figure 3. Component choice Figure 4. Moving components


Figure 5. Drawing the circuit


2.3. Devices and simulation

Two types of virtual electronics equipments are available: generators and
oscilloscopes. Generators (Fig. 6.) enable students to set up a signal in terms of
frequency, voltage and waveform. This signal will be applied to the circuit on the inputs
selected by the student. Oscilloscopes (Fig. 7.) enable students to view circuits’ outputs.
That is to say: “simulation results”. For each oscilloscope two channels are available.
Students can adjust voltage and/or time scale.


3. Virtual world distribution

The server implementation is based on the DeepMatrix software (Reitmayr et.al.,
1998) from GEOMETREK. This software enables users to enter 3D websites where they
can interact with other users and objects. DeepMatrix implements client-server
University of Bucharest and Ovidius University of Constanta

38
architecture. On the server side, all messages are broadcasted in the same order to all
clients. We have refined the proposed implementation from GEOMETREK, by
introducing a filtering and pseudo dead reckoning mechanism (Singhal and Zyda, 1999)
that permit a more friendly and flexible connection of users.


Figure 6. Generator Figure 7. Oscilloscope

Clients are Java applets running in a HTML Browser. The communication between
VRML world and client applet is made by use of External Authoring Interface (EAI)
(Fig. 8). On the client side the EAI permits to achieve complex tasks by connecting the
VRML Web Browser plug-in with a Java applet within the same web page.


Figure 8. Distributed Virtual Lab. software architecture

EAI enables a two-way communication between the Java applet and the
plug-in. The Java applet loads VRML content into the plug-in and adds avatar
representation to the virtual world.
The avatars were designed on an approximate-body approach (Capin et.al., 1999),
which provides frequently position and orientation information to remote hosts, taking
into account a minimal set of joint points. The plug-in updates the Java applet about users
position and orientation in the virtual world.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

39
DeepMatrix offers network data structures, which enable clients to share
data or to communicate together.
Concerning the Distributed Virtual Lab., VRML and Java code of each
client are similar (Fig. 8.).
The main difficulty lies in the fact that these different clients must evolve in
the same way according to users’ actions in the different virtual worlds. That
means that we have to know at the level of each java application if an event is
local (a local user action) or if it is an update from network (another user action).
For example, when a user is changing the value of a resistor we have to change
resistor’s colours in the virtual world of this user and broadcast the new value of this
resistor on the network. If the value of the resistor is changing because of a network
update we just have to change the colours of this resistor. We do not have to broadcast
anything else.
Another problem that is not solved by deepmatrix is the dynamic insertion of
VRML code into a virtual world.
For example, when students proceed to virtual welding a line is inserted into the
virtual world. If a new client connect to the virtual world, it is necessary to know if there
was any welding before its connection. The same problem arises when a user requests a
simulation. Simulation results are drawn on the screen of the virtual oscilloscope by
means of VRML code, which is automatically generated and inserted into the virtual world.
Such data are saved into a file on the server side. This file keeps a trace of the state
of the virtual world. This mechanism enables new clients to join old clients and share the
same state.


4. Virtual tutor behaviour

4.1. Describing Virtual Tutor behaviour

The first step is to find or to define a tool in order to describe the behaviour of the
Virtual Tutor.
This behaviour has to be designed by a teacher. The challenge is to provide a tool that
is easy to use and easy to understand by users who are not specialized in computer sciences.
This tool must also be “content independent”. That is to say that this tool must not be
especially dedicated to the monitoring of practical activities in the field of electronics. The
developed approach of virtual tutoring should be reused in various cases of practical activities.
Concept maps are widely used to describe experts’ knowledge from various
domains, for example in the field of electronics (Coffey et.al., 2003) or medicine
(Michael et.al.). They can also be used to help students to integrate new concepts
(Fernando Vega-Riveros et.al., 1998) even more they have been adapted with preschool
children who can’t read yet (Figueiredo et.al., 2004).
Concept maps are graphs that connect nodes with arcs. Nodes represent concepts
and arcs represent relationships between nodes. It is an intuitive and visual representation
University of Bucharest and Ovidius University of Constanta

40
technique that seems to have “more computational efficiency” than any other forms of
knowledge representation (Kremer, 1994).
According to the fact that in our Virtual Lab. most of behaviours have been
developed in java language, we are naturally guided in choosing the same tool as an
implementation language for the Virtual Tutor.
The behaviour of the Virtual Tutor is implemented by means of JESS (Java Expert
System Shell). JESS is a Java implementation based on Clips (JESS). On the one hand,
JESS has been used by other authors in order to control a virtual tutoring system and an
architecture has been proposed in order to structure JESS rules (Gutl and Pivec, 2002).
On the other hand, concept maps have also been used to formalize JESS rules
(Ciffey et.al., 2003).
But the originality of our work is to propose a generic approach (a content independent
approach) that would enable the automatic generation of a Knowledge Base from a Concept Map.


Figure 9. From Concept Map to Virtual Tutor behaviour

The different steps of our approach are showed Fig. 9:
1. Teachers design a Concept Map that represents the behaviour of the Virtual
Tutor. Of course, this design must fit with the exercise that students have to achieve.
Teachers are using CmapTools (CmapTools).
2. Rules are extracted from the Concept Map in order to feed JESS Knowledge Base.
3. The Virtual Tutor is a VRML avatar that will interact with students according
to their actions in the virtual world. JESS takes in charge the control of the avatar.


4.2. Translating a Concept Map into rules

According to the fact that:
“Concept maps are not computational unless they have an associated semantics.
That is, the maps' node and link types and their interconnections must be constrained to
allow for computer support.”(Kremer, 1994)
The 3
rd
International Conference on Virtual Learning, ICVL 2008

41
we have defined a basic semantics in order to be able to compute a Concept Map:
− Maps’ nodes are predicates or actions.
− Link type is unique and the meaning of the link is “implies”.
An example of such a map is given Fig. 10.



Figure 10. A map fragment on Operational Amplifier exercise

CmapTools enables the generation of an XML file describing the map. This XML
file contains information concerning the topology and the semantics of the map (Fig. 11).
We use semantics data from the XML file in order to generate rules as following (Fig. 12):
1. Set up a 2D array with Linking-Phrases as row and Concepts as columns.
2. Assign weights to each connection.
Weight = –1 if the connection goes from concept to linking-phrase.
Weight = +1 if the connection goes from linking-phrase to concept.
3. Extract a rule from each row.
4. Write the rule in XML format for JESS: JESSML Language (JESS).



Figure 11. XML fragment on Operational Amplifier exercise
University of Bucharest and Ovidius University of Constanta

42


Figure 12. From Concept Map to rules

This method enables us to distinguish predicates and actions from the set of
concepts. If there is a weight equal to -1 in a column that means this concept is a
predicate. Otherwise it is an action.


5. Virtual tutor implementation

The Virtual Tutor is represented in the virtual world by means of an avatar, which
is not connected to any users. This avatar is controlled by JESS on the server side (Fig.
13). Thus all clients that share the same virtual world are sharing the same Virtual Tutor.



Figure 13. Virtual Tutor implementation
The 3
rd
International Conference on Virtual Learning, ICVL 2008

43
DeepMatrix offers network data structures, which enable clients to share data or to
communicate together. DeepMatrix collects users’ interactions. Events that are relevant to
JESS rules are provided to JESS. Then JESS inference engine fires rules in order to select
Virtual Tutor actions.

Virtual Tutor actions are both text messages and sentences that are stored on server
side by means of mp3 files. When a comment has to be provided to students the Virtual
Tutor speaks to students and, in the same time, a text message is broadcast to all students.


6. Conclusions and future works

This implementation of the Virtual Tutor has been experimented in the framework
of an exercise dedicated to Operational Amplifier.
On a pedagogical point of view, it is really easy for a teacher to create a Concept
Map and to generate rules for JESS. Such an approach could be used in other context.
But the integration of JESS to the Virtual Lab., on the server side, requires some
hand works in order to link predicates to events collected by DeepMatrix. This point
should be improved by means of a dictionary of events. The teacher could use this
dictionary of events during the design phase of the Concept Map.
Semantics of our Concept Map should also be improved by means of new relations
associated to linking-phrases. The same method could be used in order to generate rules
for each type of relation.
On a technical point of view, DeepMatrix enables the use of avatar gestures.
Experiments will help us to design and implement avatar gestures according to end-users’
needs. This would help us to improve Virtual Tutor behaviours (Popovici, et.al, 2003).
We are also working on the integration of a speech synthesis module from
MBROLA Project (MBROLA). This module will enable the Virtual Tutor to speak
without needing any pre-recorded mp3 file.



REFERENCES


ANDERSON, R. E. (1992), Social impacts of computing: Codes of professional ethics. Social Science
Computing Review 10, 2, 453-469.
CAPIN, T. K., PANDZIC, I. S., MAGNENAT-THALMANN, N., THALMANN, D. (1999), Avatars
in Networked Virtual Environments, John Wiley & Sons, ISBN: 0-471- 98863-4.
COFFEY J. W., A. J. CAÑAS, T. REICHHERZER, G. HILL, N. SURI, R. CARFF, T. MITROVICH &
D. EBERLE (2003), Knowledge Modeling and the Creation of El-Tech: A Performance Support
and Training System for Electronic Technicians, Expert Systems with Applications, 25(4).
CmapTools, official Homepage, http://cmap.ihmc.us/
FERNANDO VEGA-RIVEROS, J., GLORIA PATRICIA MARCIALES-VIVAS, MAURICIO MARTÍNEZ-
MELO (1998), Concept Maps in Engineering Education: A Case Study, Global J. of Engng.
Educ., Vol. 2, No. 1.
University of Bucharest and Ovidius University of Constanta

44
FIGUEIREDO, M., LOPES, A. S., FIRMINO, R., SALOMÉ DE SOUSA (2004), “Things we
know about the cow”: Concept mapping in a preschool setting, Proc. of the First Int.
Conference on Concept Mapping, Pamplona, Spain.
GERVAL, J-P., LE RU, Y. (2006), VELab: A Virtual Lab for Electronics Virtual Experiments,
Advanced Technology for Learning, Volume 3, Issue 2, ACTA Press.
REITMAYR, G., CARROLL, S., REITMEYER, A., WAGNER, M. G. (1998), DeepMatrix – An
Open Technology Based Virtual Environment System, White Paper, October 30.
GÜTL, CH., PIVEC M. (2002), Virtual Tutor, Proc. of ED-MEDIA 2002, Denver, USA, 668-672.
JESS: The Java Expert System Shell, official Homepage, http://herzberg.ca.sandia.gov/
KREMER, R. (1994), Concept Mapping: Informal to Formal, ICCS'94, Proceedings of the
International Conference on Conceptual Structures, University of Maryland.
MBROLA, official Homepage, http://tcts.fpms.ac.be/synthesis/mbrola.html
MICHAEL J., ROVICK A., GLASS M., ZHOU Y. and EVENS M., Learning from a Computer Tutor
with Natural Language Capabilities, Interactive Learning Environments, 11(3): 233–262.
POPOVICI, D. M., SERBANATI, L. D., GERVAL, J. P. (2003), Virtual Perception Based Agents
in Virtual Theater, Technologies for Interactive Digital Storytelling and Entertainment,
TIDSE 2003, Darmstadt, Germany, march 24-26.
SINGHAL, S., ZYDA, M. (1999), Networked Virtual Environments, Addison-Wesley, ISBN:
0-201-32557- 8.
VRML Standard Version 2.0, ISO/IEC CD 14772, 1996, http://vrml.org/VRML2.0/

The 3
rd
International Conference on Virtual Learning, ICVL 2008

45
Simulation and Training with Haptic Feedback – A Review

Simona Clapan
1
, Felix G. Hamza-Lup
1


(1) Computer Science, Armstrong Atlantic State University
Savannah, GA 31419, USA
E-mail: Felix.Hamza-Lup@armstrong.edu


Abstract
Recent advances in haptic technology have broadened the applicability spectrum of
haptic devices and the potential of prototype development for commerce. This article
provides a review of the available haptic technologies and associated hardware/
software characteristics. We compare haptic devices from the hardware perspective.
We present the main features of existing haptic APIs as well as the trend in haptic
applications development. We examine several case studies to demonstrate the
effectiveness of haptic devices.

Keywords: haptic devices, virtual reality, simulation and training


1. Introduction

The word “haptics” derives from the Greek haptesthai, meaning “to touch” (Wall, 2004).
Haptics is the science enabling tactile sensation in computer applications for simulation and
training purposes. The user can receive three types of touch sensations through a haptic device:
force feedback, tactile feedback, and proprioception (from latin “proprius”, meaning “one's
own” and perception, the sense of the relative position of neighboring parts of our body).
Haptic devices apply small forces through a mechanical linkage (e.g. a stylus in the
user’s hand) (Lamoureux, 2005). Devices such as the haptic glove (Sensable
Technologies) allow the user to feel the shape and form of virtual objects, while others,
such as the Screen Rover (www.abledata.com), enable visually impaired users to access
computers almost as easily as users without visual impairments.
Our presentation is organized as follows. In section 2 we categorize haptic
applications based on their application domain. In section 3 we present a brief history of
haptic research. Sections 4 and 5 examine haptic devices and their characteristics. Section
6 explores several Application Programming Interfaces (APIs), and in section 7 we
investigate the effectiveness of haptic augmentation through several case studies.


2. Application Domains

The rapid growth of academic interest in haptic systems is stimulated by the
decreased cost of haptic hardware and the growing interest in haptic applications in the
private sector. Several haptic application domains follow.
University of Bucharest and Ovidius University of Constanta

46
2.1. General Education

Research in psychology proves that students have different styles of learning, based
on their cognitive development and abilities. Many learners understand and memorize
better when movement and touch are involved. Focused only on visual and auditory
learning, the traditional school can be inefficient for this category of students. The classic
method of teaching can be defective even for the visual and auditory learners as they
often memorize the phenomenon or process without understanding its underlying
mechanisms. Students can have a deeper understanding of the concepts when haptic
feedback is incorporated into the learning material.
The HaptEK16 simulator (Hamza-Lup, 2008) facilitates student understanding of
difficult concepts (e.g. hydraulics) and has the potential to augment or replace traditional
laboratory instruction with an interactive interface offering enhanced motivation,
retention and intellectual stimulation.


2.2. Medicine

One of the most active application domains for haptics in medicine is laparoscopic
surgery training. Additionally, surgeons at remote locations may use haptic applications
to practice surgical procedures. Several research groups worldwide currently have
surgical simulation applications. In one demonstration a “surgeon” located in Australia
guided a ”trainee” in Sweden in an operation to remove the gall bladder, using an Internet
link between Australia and Sweden (Satava, 1998).
The advancements in medical modeling and Virtual Reality enable medical training
in a safer and more cost-efficient manner. A study by Chui (Chui, 2006) analyzes a
surgical simulator for training students to perform spinal cement vertebroplasty. In this
biomechanical model a haptic device is employed to capture the movement of the user’s
hand, and the Cybergrasp™ device provides force feedback to the user’s finger during the
insertion of the needle into the bone. The Haptic Cow simulator (Baille, 2005) is another
haptic application designed for veterinary students performing fertility examinations.
During training, the students palpate virtual internal organs via a haptic device that is
positioned inside a fibreglass half-cow model.


2.3. Assistance for Visually Impaired

Haptics-enabled systems can aid blind and visually impaired users at using
computers or playing games (Brewster, 2001; Basdogan and Ho, 2002). For instance, Yu
et al (Yu, 2003) developed a low-cost web-based tool which can be used by blind people
to design virtual graphs without the help of a sighted person. The automatic graph
generation works like the graph-plotting tool in Microsoft Excel that plots a graph
according to the selected data. Based on the data inserted by the user, the tool renders a
graph on the computer screen. Blind users can then explore the graph through Logitech’s
The 3
rd
International Conference on Virtual Learning, ICVL 2008

47
WingMan Force Feedback mouse with audio feedback. The interactive drawing gives
blind users the opportunity to draw graphs manually.


2.4. Military Simulations and Training

Haptic-enabled VR simulations across a network allow people in different locations
to participate in military training exercises (Gun and Mettenmeyer, 2002). At the Army's
National Automotive Center, the Simulation Throughout the Life Cycle program used
haptics to test military ground vehicles under simulated battlefield conditions. For
example, they simulated an environment where workers at remote locations can
collaborate in reconfiguring a vehicle chassis with different weapons using instrumented
force-feedback gloves to manipulate the 3D components.
Haptic applications can be used to safely train aircraft and other complicated
machinery operators. Flight training simulators are safer when teaching potentially
dangerous tasks, such as taxiing down a runway (Menéndez and Bernard, 2001).


2.5. Architecture and Graphic Arts

Haptics may be applied to designing virtual art exhibits, concert rooms, museums,
and even individual or co-operative virtual sculpturing projects across the Internet
(Brewster, 2001; Handshake VR News, 2004).
Novint™ Technologies developed an architectural walkthrough for Sandia National
Laboratories that allows users to load detailed architectural models and explore their
design using Novint’s e-Touch technology. Haptic technology allows users to receive
haptic feedback while feeling the digital models, or picking up and placing objects such
as chairs.


2.6. Entertainment

Haptics naturally fits in video games and simulators by enabling the user to feel and
manipulate virtual solids, fluids, tools, and avatars (Handshake VR News, 2004). One
example is a stock XBox controller (Basdogan and Ho, 2002) powered by Immersion’s
force feedback technology. Players of “Star Wars” game have the opportunity to
experience a heavy recoil effect when firing a rocket launcher and the rapid-fire
vibrations from a machine gun.


3. Haptics Research

Haptic research originates with the work of Heinrich Weber (Prytherch, 2002), a
19
th
century professor at the University of Leipzig. In 1987 Lederman and Klatzky
(Klatzky, 1985), summarized four basic procedures for haptic exploration, each one
eliciting a different set of object characteristics:

University of Bucharest and Ovidius University of Constanta

48
• lateral motion (stroking) provides information about the surface texture of the object;
• pressure gives information about how firm the material is;
• contour following elicits information on the form of the object;
• enclosure reflects the volume of the object.

The development of several haptic devices in the early 1990s facilitated important
experiments that involve human tactile perception, and improved the understanding of
haptic human-computer interaction.
The increasing number of researchers in the haptics domain in the late 1990s
contributed to the appearance of a specialized Internet magazine. Haptics-e published
haptics-related technical discussions and articles. Since the foundation of Haptics-e (2000)
and Haptics International Society (2003), numerous conferences, symposiums, and
publications were organized, indicating the expansion of the haptics research community.


4. Haptic Devices and Hardware Characteristics

In this section we present the most novel haptic devices and we categorize the
hardware device characteristics by comparing information from various manufacturers.
As we mentioned earlier, blind users can then explore the graphs through
Logitech’s WingMan Force Feedback mouse (figure 1) with audio feedback. Another
successful initiative pursued by SensAble Technologies is their line of PHANTOM®
devices. PHANTOM® Omni™ (figure 2a) is a six-degree-of-freedom portable device
with a compact footprint and a removable rubber stylus. An alternative tool is the
PHANTOM® Desktop™ (figure 2b), which is similar to the PHANTOM® Omni™, but
provides better precision positioning control and higher fidelity force feedback output.
The Mimic Mantis has a different design compared to other haptic devices: this
tension-based device incorporates an on-board processor for faster computation of forces,
allowing the haptics software to be embedded directly into the device. The user interacts
with the system through an integrated keyboard and a two-button grip, which can be
changed to satisfy the application requirements.
The wireless CyberGlove® II from Immersion Corporation (figure 3) is a fully
instrumented glove that provides up to 22 high-accuracy joint-angle measurements. It
uses proprietary resistive bend-sensing technology to accurately transform hand and
finger motions into real-time digital joint-angle data. Each of the incorporated sensors is
extremely thin and flexible, being virtually undetectable in the lightweight elastic glove.



Figure 1. Logitech’s
WingMan™ mouse
Figure 2a.
PHANTOM® Omni™
Figure 2b.
PHANTOM® Desktop™
Figure 3.
Cyber Glove II
The 3
rd
International Conference on Virtual Learning, ICVL 2008

49
Novint Technologies, Inc. introduces a 3D game controller haptic device, Novint
Falcon, which (figure 4) enables users to control a game in three dimensions. The device
has three arms that are gathered in a handgrip with programmable buttons. The position
and the forces rendered by each arm are updated 1000 times per second to create a real
life experience for the user.
Having developed a set of prototypes since 1997, Carnegie Mellon University
proposed in March 2008 an innovative haptic device based on magnets. The device
(figure 5), built into a bowl-shaped cavity in a desk, includes a levitating bar that, grasped
by the user, makes the magnets exert force on the bar. Missing the mechanical linkages,
the system responds instantly due to no latencies from mechanical force-feedback. The
moving part of the device, which responds and exerts actions in six degrees of freedom,
exhibits relatively high stiffness (25 N/mm at 1500 Hz) and allows appliance of forces up
to 55N. The perception of very smooth movements of the bar (up to 5-10 microns)
permits the feel of differences between textures and subtle effects of friction. The
disadvantage of the system is the limited range of motion for the joystick: 25 mm in
translation and 15-20 degrees in rotation.



Figure 4. Novint Falcon Figure 5. Magnetic levitation (maglev)

The following characteristics, also known as performance measures, are common to
all haptic devices (Wall, 2004):
• Degrees of Freedom (DOF) represent the set of independent displacements that
specify completely the position of the body or system.
• Workspace refers to the area within which the joints of the device will permit the
operator’s motion.
• Position resolution is the minimum detectable change in position possible within
the workspace.
• Continuous force is the maximum force that the controller can exert over an
extended period of time.
• Maximum force/torque is the maximum possible output of the device,
determined by such factors as the power of the actuators and the efficiency of any
gearing systems. Unlike continuous force, maximum force needs to be exerted
only over a short period of time (e.g., a few milliseconds).
• Maximum stiffness of virtual surfaces depends on the peak force/torque, but is
also related to the dynamic behaviour of the device, sensor resolution, and the
sampling period of the controlling computer.
• System latency measures the time passed between the moment of changing the
controller’s position and the moment when a resultant force can be calculated and
University of Bucharest and Ovidius University of Constanta

50
rendered by the device. Latency includes computation by the computer and
therefore depends on the speed of the computer as well as the speed of the device.
• Haptic update rate is the inverse of system latency, measured in Hz.
• Inertia is the perceived mass of the device when it is in use. This should be as
low as possible to minimize the impact of the device controller on rendered forces.


5. Haptic APIs

Several APIs have evolved for the development of haptic applications. They
include SensAble OpenHaptics Toolkit, Reachin API, Immersion Corporation’s API for
automotive, and Sense Graphics’ H3D API.
SensAble OpenHaptics Toolkit (www.sensable.com) enables software developers
to add haptics and 3D navigation to a wide range of applications, from games and
entertainment to simulation and visualization. The toolkit is familiar to graphics
programmers because it is designed after the OpenGL API.
Reachin API (www.reachin.se) is a modern development platform that enables the
development of sophisticated haptic 3D applications in the user's programming language
of choice, such as C++, Python, or VRML (Virtual Reality Modeling Language). The
API provides a base of pre-written code that allows for easy and rapid development of
applications that target specific needs of the user. UK Haptics, a newly established
medical software development company, agreed to use Reachin API as the core haptic
technology platform for their Virtual Veins application. Virtual Veins is a medical
simulation package for training medical staff in catheter insertion.
The Immersion API (www.immersion.com) is a software library for creating and
assigning haptic effects to interact with haptic devices such as rotary controllers. It
provides the code necessary for developers to design and incorporate haptic effects into
their applications. Leading auto manufacturer, BMW, has licensed Immersion's
TouchSense technology to create the automotive industry's first intuitive information and
control system called iDrive. The iDrive features a single control dial mounted on the
central console, which allows a driver to have instant and total control of every comfort
element in the car through their sense of touch.
H3D API is designed mainly for users who want to develop haptics applications
from scratch, rather than for those who want to add haptics to existing applications. The
main advantage of H3D API is that it makes it easy to manage graphics and haptics
rendering. For this reason, H3D API is a vital extension to OpenHaptics. It allows users
to focus their work on the behavior of the application and ignore the issues of haptics
geometry rendering as well as synchronization of graphics and haptics. The API is also
extended with scripting capabilities, allowing the user to perform rapid prototyping using
the Python scripting language.


6. Effectiveness of Simulation and Training With Haptic Feedback

In a study Moody et al (Moody, 2002) demonstrated the effect of a force feedback
system in the training and assessment of surgeons. The PHANToM desktop unit, run on
Windows NT 4.0, was used together with a suturing simulation. After the task was
The 3
rd
International Conference on Virtual Learning, ICVL 2008

51
demonstrated and explained to each subject by the experimenter, each of the 20
participants performed two test sutures to familiarise themselves with the task and the
experimental setting. Participants were then asked to form one suture across the skin
excision, with the specifications provided by the experimenter. Results revealed that force
feedback resulted in a reduction of the time taken to complete the stitch.
Cagatay Basdogan et al. (Basdogan, 2000) have conducted experiments to study the
role of haptic feedback in performing collaborative tasks and in influencing the sense
togetherness when working with a remote partner. For this purpose they designed a
multimodal shared environment that included: one computer, two synchronized monitors,
and two PHANToM devices. The 10 participants formed two groups. In one day one of
the groups performed the task including only visual feedback, while for the other group
visual and haptic feedback was included. In each of the 15 trials the participants
collaborated with their partner to move a ring in virtual environment without touching a
wire. The results of the experiment suggest a considerable enhancement of performance
when the haptic feedback was present. The measurements also revealed that, depending
on the age, the gender, and the level of computer usage of the participants, the haptic
presence increased in some level the feeling of togetherness.
The study (Caroline, 2007) analyzes the effects of network delay on users that are
working in a collaborative environment. Thirty participants took part in the study,
performing the experiment in pairs. For observing the differences between the visual and
haptic latency, the experimental task consisted of two parts: in the first part the users,
positioned in a simple environment at some distance one from another, had to get close to
one another relying on visual feedback; in the second part they had to move to a target,
without loosing contact, relying on haptic feedback. Pairs of participants performed 12
experimental sessions with random level of latency added for every trial. The negative
effects of the latency were slowed movement and an increased number of errors.
Virtual Haptic Back (VHB) Project (Williams, 2006) develops a series of haptic
simulations of the human body parts, such as somatic dysfunctions, to help students learn
the palpatory techniques. The project includes passive and active methods of study. The
application has a multistage structure. The path and the movements of the expert
performing the palpatory technique are recorded using PHANToM playback capabilities.
The first stage allows students to follow the expert’s path, with no haptic feedback
incorporated. In the second phase the haptic feedback is involved, and the student has to
actively follow the correct path via visual cues. The results of the experiment show that
the users from both groups improved their technique during the trials; however, the students
from the group trained with passive trials performed significantly better then the other group.
As confirmed in the above case studies, the use of haptic feedback in simulation
and training seems to improve the user’s experience and efficiency of performed
procedures at a cost of a more complex system.


7. Conclusion

In this paper we provided a review of the available haptic technologies and associated
hardware/software characteristics. Haptics is a fast-growing field with serious potential and a
multitude of applications in entertainment, medicine, military and other fields. Several haptic APIs,
stand out and enable faster development of haptic applications for simulation and training purposes.
University of Bucharest and Ovidius University of Constanta

52
The efficiency of simulation and training with haptic feedback is demonstrated for
several application domains. However designing a training tool with haptic feedback
increases the complexity and the real-time processing requirements of the application.
Significant progress has been made since the inception of the technology, and we
believe that even more innovative haptic applications will be seen in the future.



REFERENCES


BAILLIE, S., CROSSAN, A., BREWSTER, S., MELLOR, D., and REID, S. (2005), “Validation of
a Bovine Rectal Palpation Simulator for Training Veterinary Students”, in Medicine Meets
Virtual Reality 13: The Magical Next Becomes the Medical Now, Vol. 111, pp.33-36.
BASDOGAN, C., HO, C., SRINIVASAN, M. A, and SLATER, M. (2000), “An experimental
study on the role of touch in shared virtual environments”, in the ACM Transactions of
Computer-Human. Interaction, Vol. 7(4) , pp. 443-460.
BASDOGAN, C., and HO, C. (2002), “Principles of Haptic Rendering for Virtual Environments”
(network.ku.edu.tr/~cbasdogan/Tutorials/haptic_tutorial.html).
BREWSTER, S. A. (2001), “The Impact of Haptic 'Touching' Technology on Cultural Applications”,
in the Proceedings of EVA2001, Glasgow, UK, pp. 1-14.
CHUI, C.-K., ONG, J. S. K., LIAN, Z.-Y., WANG, Z. and TEO, J. (2006), “Haptics in Computer-Mediated
Simulation: Training in Vertebroplasty Surgery”. Simulation and Gaming, Vol. 37, pp. 438-451.
GUNN, C. and METTENMEYER, A. (2002), “Virtual Surgery across the World”, CSIRO Media
Release – Ref 2002/224 – Nov.13, 2002.
HAMZA-LUP, F. G. and ADAMS, M. (2008), "Feel the Pressure: e-Learning System with Haptic
Feedback", The 16
th
Symposium on Haptic Interfaces for Virtual Environments and
Teleoperator Systems, March 13-14, Reno, Nevada.
HANDSHAKE V R News (2004), “Telehaptics for Training and Command and Control”, The
MSIAC’s M&S Journal Online, Vol. 5.
JAY, C., GLENCROSS, M., and HUBBOLD, R. (2007), “Modeling the effects of delayed haptic
and visual feedback in a collaborative virtual environment”, in the ACM Transactions of
Computer-Human Interactions, Vol. 14 (2).
KLATZKY, R. L., LEDERMAN, J. and METZGER, V. A, (1985), “Identifying objects by touch:
An expert system”, Perception and Psychophysics, Vol. 37, pp. 299-302.
LAMOUREUX, M. (2005), “Making Sense of Human/Machine Interface Controls. Unmanned
Systems”, (www.ultra-msi.com/company/news/2005_12_01.pdf).
MENÉNDEZ, R. G. and BERNARD, J. E. (2001) “Advancing the State of the Art in Flight
Simulation via the Use of Synthetic Environments”, Iowa Space Grant Consortium.
MOODY, L., BABER, C. and ARVANITIS, T. N, (2002), “The Role of Haptic Feedback in the
Training and Assessment of Surgeons”, in proceedings of Eurohaptics 2001, Birmingham,
UK. University of Birmingham, pp. 170-173.
PRYTHERCH, D. (2002), “Weber, Katz and Beyond: An Introduction to Psychological Studies of
Touch and the Implications for an Understanding of Artists' Making and Thinking
Processes”, in Research Issues Art, Design and Media, Vol.2.
SATAVA, R. M. and JONES, S. B. (1998), “Current and Future Applications of Virtual Reality
for Medicine”, in Proceedings of the IEEE, Vol. 86, pp. 484-489.
YU, W, KANGAS, K and BREWSTER, S. A. (2003), “Web-based Haptic Applications to Allow Blind
People to Create Virtual Graphs”, in Proceedings of the 11
th
Haptic Symposium Los Angeles, CA.
WALL, S. (2004), “An Investigation of Temporal and Spatial Limitations of Haptic Interfaces”.
Department of Cybernetics, vol. Ph.D. Reading: University of Reading.
WILLIAMS R. M. II, SRIVASTAVA, M, CONATSER, R. R, and HOWELL, J. N., (2004)
“Implementation and Evaluation of Haptic Playback System”, http://www.haptics-e.org, 3.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

53
INTERGEO – Interoperable Interactive
Geometry for Europe

Christian Mercat
1
, Paul Libbrecht
2
,
Sophie SouryLavergne
3
, Jana Trgalova
3


(1) I3M, LIRMM, Univ. Montpellier 2, France
(2) German Research Center for Artif. Intel. (DFKI), Saarbrücken, Germany
(3) National Institute for Pedagogical Research (INRP), Lyon, France

Abstract
Intergeo (http://inter2geo.eu/) is an eContent+ European project dedicated to the
sharing of interactive geometry constructions. It will enable teachers and pupils all
over Europe to participate from experiences made by pioneers in the field of
interactive geometry as a tool for teaching, learning, and research. Educational
contents that were hard to access are made available in a common interoperable
format. Tagged with relevant topics and competency based metadata and
categorised according to curricula, they will be searchable and easily (re)usable by
everyone. It will impact the value chain in eLearning, by providing building blocks
of quality controlled, semantically enriched interactive educational content, on all
levels from K12 to university, for classrooms, online courses, or integrated digital
education systems. This project will help a multicultural community, built around
interoperable quality controlled eLearning standards, to emerge and sustain itself
with a wider audience than the present days niche of dedicated experts, which
would not happen by market forces alone.


1. Introduction

The last decade has seen a bloom in tools that allow teachers to enrich their
teaching with interactive data, whether in face to face or distant mode. This wealth has its
drawbacks and teachers need support to navigate through this diversity: which software
should I use, where can I find resources, will this resource work for my class? Indeed,
apart from pioneer work by dedicated teachers, the actual practices in the classroom have
not evolved much. The reasons are manifold. Here are the three main ones:
– All the communities that have grown around the different technical solutions and
software available have produced resources that they share in one way or another. They
have all thought about their practice and produced different approaches. Currently these
cannot be merged, because the data they produce is scattered, both physically and
semantically. The resources need to be centrally visible and exchangeable.
– As well as being difficult to find and analyse, the resources are usually diverse in
quality and relevance to a specific need. Teachers are unsure in which situation a given resource,
even if apparently interesting, could actually be used, and whether it adds pedagogical
value to the learning experience. They wait for a bolder colleague to report on her
attempt. The resources need to be tested, and published reports need to reflect these tests.
University of Bucharest and Ovidius University of Constanta


54
– Mastering a piece of software is time-consuming, and very few teachers grow to
become power-users of their tool. The resources need to be easy to use, share and adapt,
in spite of software choices.
In order to solve these issues at least for one specific subject, interactive geometry,
we propose to centralize educational resources from this field on the Intergeo web
platform. All resources will have clear Intellectual Property Rights, promoting open
licences. And they will be there in an interoperable file format we are going to create,
based on OpenMath [3]. This format will be supported by the most common software
programs for interactive geometry, so teachers can keep on using their own.
We proceed as follows. First, we defined an internationalized ontology describing
our field of interest. Second, we annotated curricula of various countries with items from
this ontology. Third, we let the users annotate their resources and browse through existing
resources using the nodes of the ontology. Their use is reported through an online quality
questionnaire that helps ranking the resources and identifying improvement axis.


1.1. Outline

The introduction continues with explaining what interactive geometry is and who
we are. Section 2 describes the aim of the project. Section 3 presents the metadata based
on a multilingual ontology used for both representing the various European curricula,
educational levels, and the competencies attached to the resources. Section 4 explains
how this metadata allows to search and access the content, how queries are processed, by
typing and explicitly selecting competencies and topics or by pointing in a curriculum or
a textbook. Then, in section 5 we describe how the evaluation of the quality of the
resources will be performed and used. The paper ends with a conclusion Sec. 6.


1.2. What is Interactive Geometry?

The Intergeo project is driven by European leaders in interactive geometry
software. We are going to explain what is understood by interactive or dynamic
geometry, a way of doing geometry which is required of math and science teachers more
and more often. Interactive geometry allows for the manipulation and the visualization of
a construction (a figure) on a computer. The construction depends on some free
parameters, like the position of one or several control points. The user manipulates the
figure through the keyboard, the mouse or a tracking device, by changing one or more of
these free parameters. The construction then changes accordingly.
Let us give a simple example. One constructs in a dynamic interactive geometry
system a triangle ABC with two perpendicular bisectors of two sides, then the
intersecting point O of these two perpendicular bisectors. The third perpendicular bisector
is then constructed and seems to pass through the point O. In a dynamic geometry system,
it is possible to drag any vertex of triangle ABC and although the shape of triangle ABC
is changing, the third perpendicular bisector is always passing through the point O as
depicted in figure 1.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

55


Figure 1. Perpendicular bisectors

The property of three perpendicular bisectors intersecting in a common point
appears as an invariant when varying the triangle ABC. Being able to move screen
objects around in space (and so over time) can add significantly to the user’s sense of the
underlying concept as an object not just in itself but a something invariant amidst change.
Interactive geometry is intended to manipulate scientific data relying on a
hierarchical construction. The figure encodes not only the graphical illustration (a curve
here, a picture there) but also the relations between the different entities that are drawn.
Of course, the main entities and relations in interactive geometry are of geometrical type.
You will find triangles, circles, lines and points, barycentres, tangents, secants with given
angles and distances [14]. But it is much more general than antique Greek geometry –
you can have functions, derivatives, colours, random variables, all sorts of constructs that
allow you to visualise and manipulate concepts that arise in all sorts of contexts, inside
mathematics as well as outside [1, 5, 7].


1.3. Who we are

Intergeo (http://inter2geo.eu) is a project funded by the European community under
the eContent+ programme. It begun in October 2007 and will last for three years.
Our consortium brings together partners from six different European countries. The
geometry software that has been developed inside the consortium, however, covers
almost all languages of the twenty seven countries of the European Union (and more).
1. University of Education Schwäbisch Gmünd, Germany [Cinderella]
2. Université Montpellier 2, Sésamath association, GNU Edu, France [Geoplan/
Geospace/Tracenpoche]
3. German Research Center for Artificial Intelligence (DFKI), Saarbrücken, Germany
4. Cabrilog SAS, France [Cabri II Plus/Cabri Junior/Cabri 3D]
5. University of Bayreuth, Germany [GEONExT]
6. Université du Luxembourg [Geogebra]
7. University of Cantabria, Santander, Spain
University of Bucharest and Ovidius University of Constanta


56
8. TU Eindhoven, Netherlands
9. Maths for More, Spain [WIRIS]
10. Jihoceská univerzita, Ceské Budejovice, Czech republic


2. Objectives

The InterGeo project intends to ease the access to and thus enable the use and reuse
of eLearning content based on interactive geometry tools.
Development, generalisation and improvement of geometry content suffers from a
scattering of the available software and resources and a lack of quality control.
The answers we propose are:
1. Interoperability and metadata: We define and agree on a common description of
metadata and basic structure of educational interactive geometry resources, through an
ontology definition and an open file format. The resources will be easier to find, identify
and use. The common metadata and interoperable OpenMath XML specification for
describing figures in interactive geometry will permit a teacher to find, trust, open and
adapt the available resources, according to licenses. The specifications will comply with
the current standards for learning objects in order to ensure future use and sustainability.
2. Content: We will provide a wealth of own content to jump-start the exchange and
evaluation of content. Due to the achieved interoperability, user communities from
different countries have a chance to actively work together towards a better learning
experience, although they have different general conditions, different backgrounds and
pedagogical concepts.
3. Quality Assessment and User Reviews: We will help to build a common basis of
quality standards that enables users to assess the quality of content with respect to
teaching situations. To this end, we will build an equality framework able to produce,
through an assessment protocol, metadata asserting the quality, the adequacy and the
intended pedagogical use of a given resource in a given cultural context.
Great expertise in eLearning and eQuality assessment has been gathered in recent
years, in particular in some European networks such as UNFOLD or MINERVA eQuality
projects. Our project builds on the reflection, quality specifications and good practices
founded as a result of such projects.
Interactive geometry is a playing ground for multilingual share of educational
resources because its very objects are abstract and visual. Of course, pedagogical
documents have to be translated and adapted for every specific Community of Practice,
but this is not the major obstacle for a user once the genuine interactive geometry content
(as provided by the consortium) is identified.
Learning Object Repositories (LORs) are a traditional platform type to propose
sharing of learning objects. This generality implies shallow annotation standards such as
LOM, that failed at providing efficient retrieval mechanisms. Our ontology based
mechanism is much more specific.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

57
3. Interoperability and Metadata

The Intergeo project was facing the issue of cross-curriculum search which would
be fine grained because the focus on mathematics asks for more specific identification. We
give an example, review the issues and projects that addressed them and describe our solution.


3.1. A Simple Example of Cross-Curriculum Search

Consider the competency (or skill) of constructing the division of a segment in n
equal parts. This should be matched by queries using strings such as “divide in equal
parts”, “diviser en parties de même longueur”, etc. Curriculum standards, however, do not
all speak about this topic in the same way. The English curriculum only mentions the
operation of enlargement, whereas the French national program of study mentions
“connaître et utiliser dans une situation donnée les deux théorèmes suivants” and provides
the formulation of the “Théorème de Thalès” and its converse [11]. All these should
match in some way.
Mismatching across some of the curriculum boundaries is easy: In French (théorème de
Thalès) and Spanish (teorema de Tales) indicates the intercepting lines theorem. However,
Thales’ Theorem in English or in German (Satz des Thales) refers to another theorem.


3.2. Similar Projects and Approaches

Topical information in learning objects repositories is usually very broad like the
WebALT repository [8], close to a curriculum standard. Another approach is free tagging but
it needs unsustainable multi-cultural support. GNU Edu [12] provides topical information
directly within the curriculum: learning objects are tagged by skills described in a curriculum,
split into years and chapters. Skills have translated keywords to achieve cross-curriculum
search. TELOS from the LORNET research network [13] aims at complete courses and
not individual resources. England’s Curriculum Online [2], Microsoft Lesson Connection
[10] and the ExploreLearning [4] enterprise, have annotated the curricula of England and
the USA. The CALIBRATE project [18] provides annotated but too few curricula.


3.3. The GeoSkills Ontology

The basis of our approach is to enrich usual LOM like metadata [16] with a list of
mathematical competencies [9] (prerequisite or trained), topics, educational levels and
programmes which have names in many languages and which can be tagged on each
resource. These lists are arranged as an ontology so as to provide a knowledge management
tool and standards-based interoperability with guaranteed computational results.
Example of topics: Isosceles Triangle; of competencies: Calculate trigonometric
ratio; of pathways: elementary-school; of levels: Gymnasium Saarland 7te.
University of Bucharest and Ovidius University of Constanta


58
We are working on a competency editor web-based tool: it will complement the
curriki-based platform to allow: graphically browsing the competencies, topics, levels,
and their relationships (e.g. from a resource annotated with a given topic), translating,
adding or editing various names, curriculum-encoding, creating and editing competencies
and topics present in a given curriculum.


4. Content sharing

4.1. Resource model

The model for a resource stems from the work of the SFoDEM project [6] at the
genesis of the Intergeo project.
A full fledged resource in the SFoDEM form is as a collection of sheets: a learner
sheet, a teacher sheet, a technical sheet and some others. Only the learner sheet is visible
to an unidentified visitor so that learners can be directed to the resource page for an
online use of the resource.
Each sheet consists of a wiki page where the insertion of interactive geometry
constructions is done in an easy to use wiki syntax in the same way as static images. All
sheets are exported, together with the construction files, in a downloadable bundle that
can be used offline.


4.2. Resource browsing

The Intergeo platform’s main goal is to allow sharing of interactive geometric
constructions and related materials. Overall, sharing a resource is the execution of the
following roles: the author provides content to the Intergeo platform; the annotator
provides authoring, licensing, topical, and pedagogical information on it; the searcher
navigates and searches through the platform’s database to find relevant resources to use in
teaching; the curriculum encoder inputs and maintains the set of topics, competencies,
and educational contexts; the competency translator maintains their formal as well as
everyday names; the quality evaluator reports on her usage of the resource in the
classroom through an online questionnaire [17].
Topics, competencies, and contexts are addressable through URLs identifiers,
thanks to OWLdoc. A browser version can be seen from http://i2geo.net/ontologies/dev/.
We designed two means to let the users easily designate tokens (topics,
competencies, or educational contexts): by typing text or by pointing in a book.
We extend the familiar autocompletion: both for search and annotation, users can type
a few words in a text field and the autocompletion popup presents a list of matching tokens
(see Figure 2). More information and tests at http://www.activemath.org/projects/SkillsTextBox/ .
The 3
rd
International Conference on Virtual Learning, ICVL 2008

59


Figure 2. Competencies triggered by "Thales conf"

We will allow graphical browsing of curriculum standards or textbooks that users
know well. The idea is that a user can then browse through a table of content, through
pages he is graphically familiar with, and click on sections of interest. This click will
trigger the selection of the competencies and topics associated with these sections, the
search field. Although we shall mostly not be able to offer whole textbooks to browse
through, we expect it to be unproblematic to display their tables of contents.


5. Quality assessment

Quality assessment of eLearning has slowly evolved into a clear necessity. In the
Intergeo project, the quality assessment is based on user’s evaluation reports in the form
of a questionnaire to be taken by teachers to evaluate different aspects of the quality of
their planned or passed teaching experience, in order to give a ranking score to resources
and to identify directions of possible improvements.
Our methodology stems from previous projects namely the JEM (Joining
Educational Mathematics) network, the eQuality project and the IREM project SFoDEM.
The first objective of quality assessment is searchability: we want the “good” resources
to be ranked first by a search engine. The second one is reusability by improvement of
resources and their metadata through quality cycles based on users’ feedback.
In this second year 2008-2009 we will bootstrap these quality improvement cycles by
organising tests of resources in the classroom and analysing quality reports from users’ evaluations.
University of Bucharest and Ovidius University of Constanta


60
5.1. Processes

We adapted the educational model proposed in the eQuality project [15] to our
situation: Open Distance Learning provides a student with learning material from an ODL
university course. Our objective is to provide a teacher teaching content using resources
found on our web-site. The roles are somewhat shifted, as summarised in Table 1, but the
need for emulation and support, feedback and analysis are strikingly akin.

Table 1
The correspondence between eQuality and Intergeo models

e-Quality in ODL Student Teacher Institution Course Learning event
Intergeo
Teacher Author Project Resource Teaching event


5.2. Author with hats

The teacher using a resource goes around the cyclic process described in Fig. 3 and
the resource itself follows a similar cyclic process. The author of a resource has several
hats on her head to manage this cycle.



Figure 3. Teacher's workflow


5.3. Licenses

Unclear licenses are a real impediment to the use of resources found on the Internet. The
Intergeo project aims at rising the awareness of the share holders in the value chain to this
The 3
rd
International Conference on Virtual Learning, ICVL 2008

61
issue. The author chooses a license contract for the content that she intends to share. The
Intergeo project is promoting the use of open licenses that allow adaptation and reusability
such as the Creative Commons Share Alike license. According to the licence, a teacher is
encouraged to take on his own hand to improve a given resource and issue a new version
of it, taking into account the users' feedback given by the questionnaire and the forums.


6. Conclusion

The Intergeo project is reaching usability. We welcome participation from interested
users, mainly secondary math teachers to try and report on the contents on our web site
http://inter2geo.eu/. We are seeking participation of Textbooks Editors and Publishers, as
well as officials from Education Ministries in order to get competencies and up to date
curricula. The community in eLearning is welcome to enrich and use our curricula and
ontology of competencies, as well as the tools that we developed. Our ontology GeoSkills
can be tested at http://i2geo.net/ontologies/dev. The SkillsTextBox GWT project can be
enjoyed at http://ls.activemath.org/ projects/SkillsTextBox.


7. Acknowledgements

We wish to thank Odile Bénassy, Cyrille Desmoulins, Colette Laborde, Michael
Dietrich, Maxim Hendriks and Albert CreusMir for their participation and contribution to
this research. The Intergeo project is partially funded by the European Union under the
eContentPlus programme and the authors’ institutions.



REFERENCES


[1] A. Ait Ouassarah. Cabri-géomètre et systèmes dynamiques. Bulletin de l’APMEP (433):223-232, 2001.
[2] British Educational Communication and Technology Agency. Curriculum online, April 2008.
http://www.curriculumonline.gov.uk/.
[3] STEPHEN BUSWELL, OLGA CAPROTTI, DAVID CARLISLE, MIKE DEWAR, MARC
GAËTANO, and MICHAEL KOHLHASE, The openmath standard, version 2.0. Technical
report, The OpenMath Society, June 2004, http: //www.openmath.org/.
[4] ExploreLearning. Correlation of gizmos by state and textbooks, 2005, http://www.explorelearning.com.
[5] R. FALCADE, C. LABORDE, and A. MARIOTTI, Approaching functions: Cabri tools as instruments
of semiotic mediation. Educational Studies in Mathematics (66.3):317-333, 2007.
[6] D. GUIN, M. JOAB, and L. TROUCHE (eds), Conception collaborative de ressources pour
l’enseignement des mathématiques l’expérience du SFoDEM (2000-2006), Technical report,
CDROM INRP, ISBN 9782734210986 Réf. BD 151, 2008.
[7] M. HOHENWARTER, GeoGebra: Dynamische Geometrie, Algebra und Analysis für die
Schule. ComputeralgebraRundbrief (35):16-20, 2004.
[8] JOUNI KARHIMA, JUHA NURMONEN, and MATTI PAUNA, WebALT Metadata = LOM + CCD,
in Proceedings of the WebALT 2006 Conference. The WebALT project, jan 2006.
University of Bucharest and Ovidius University of Constanta


62
[9] E. MELIS, A. FAULHABER, and S. EICHELMANN, A. and NARCISS, Interoperable competencies
characterizing learning objects in mathematics. Intelligent Tutoring Systems, 5091:416-425, 2008.
[10] Microsoft. Microsoft Lesson Connection Launched At Technology + Learning Conference,
1999, http://www.microsoft.com/presspass/press/ 1999/nov99/lessonpr.mspx.
[11] Ministère de l’Education Nationale, Programmes des classes de troisième des collèges.
Bulletin Officiel de l’Education Nationale (10):108, 1998.
[12] OFSET. GNU Edu, 2008. http://gnuedu.ofset.org/.
[13] G. PAQUETTE, An ontology and a software framework for competency modeling and
management. Educational Technology and Society (10.3):1-21, 2007.
[14] J. PHILIPPE, Exploiter les logiciels de géométrie dynamique. 4 constructions géométriques
avec Géoplan. Les Dossiers de l’ingénierie éducative (54):35-37, 2006.
[15] The eQuality consortium, http://www.equalityeu.org/, 2004.
[16] The Intergeo Consortium. D2.4 metadata specification, 2008, http://www. inter2geo.eu/en/deliverables.
[17] The Intergeo Consortium. D6.1 quality assessment plan, 2008. http: //www.inter2geo.eu/en/deliverables.
[18] FRANS VAN ASCHE, Linking learning resources to curricula by using compotencies, in first
International Workshop on Learning Object Discovery and Exchange, Crete, 2007.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

63
Simulation Models for Virtual Reality Applications

Grigore Albeanu

Spiru Haret University,
13, Ion Ghica Str, RO-30045, ROMANIA
E-mail: galbeanu@gmail.com


Abstract
The paper describes some simulation models used to implement virtual reality
applications, addressing the presentation of the architecture of VR systems, VR
applications in different fields, including medicine, an introduction to simulation
techniques and a set of mathematical models for creating virtual scenes. The
material represents a significant development of the presentation given at the
workshop VRRM 2007: Virtual Reality in Rehabilitation Medicine, with details on
mathematical aspects.

Keywords: Modelling, Simulation, VR applications


1. Introduction

During last decades, many modelling methods were proposed not only for computer-
aided design, multi-queueing systems design, scientific visualization, e-learning, but also
for entertainment. Recently, more research was dedicated to modelling virtual worlds, to
model the behaviour of objects belonging to virtual environments (Dimitropoulos et al,
2008; Pasc et al, 2007; arcă et al, 2008; Jung et al, 2005; Popovici, 2005; etc), and to
simulate such a behaviour using computer graphics tools (Falcidieno & Kunii, 1993;
Hagen et al, 1993; etc), virtual reality interfaces (Fuchs & Moreau, 2003; etc.) and
languages, and augmented reality (arcă et al, 2008).
This paper describes some of architectures suitable for VR applications (the second
section) and illustrates appropriate simulation techniques (in the third section). In order to
implement such techniques, not only information technologies are required but also a
strong background in mathematical modelling. Some mathematical models based on
recent developments are described in the fourth section.
Examples from medical applications, computer-aided design, scientific-visualization,
e-learning and computer games are provided along the presentation.
The material represents a significant modification of the presentation given at the workshop
VRRM (Albeanu, 2007), with details on mathematical aspects and the current state of the art.


2. Some architectures of VR applications

Simulated VR applications can be developed for important fields, according to
(Albeanu, 2006): virtual current activities (e-learning, training in different subjects, games),
virtual “teleportation” (virtual tourism, the study of micro and nano-structures, fluid flow
University of Bucharest and Ovidius University of Constanta


64
visualization, volume visualization in medicine), virtual collaborative activities (network
based games, teleconferences, virtual communities), virtual design (CAD, architecture, fashion),
virtual management (urban management, workplace management, workstation usability,
environmental protection), virtual exhibitions (antiquities, restoration, …), and virtual events
(the study of different civilizations, old sites visiting, police investigation by replay, …).
Mainly, all applications need the participation of humans. Only some of them are
off-line simulations. Not only real humans, but also virtual characters will be parts of
some VR applications. This is why human modelling and animation is an important topic.
Hence such a conceptual model includes a human model and an environment model. Of
course, an interaction model will be also considered.



Some architecture of a VR application is based on hierarchical decision graph evaluated
repeatedly during simulation or normal running. Other VR applications use state machine
transitions (different kind of automata, including cellular), cybernetic architectures based on

Human

Virtual Environment

Actuators

Transductors /
Sensors

Sensorial
Interface

Actorical
Interface
Action/Body
Commands
Action/Tools
Information
Signals
Information
Figure 1. Virtual laboratory

The 3
rd
International Conference on Virtual Learning, ICVL 2008

65
feedback, etc. However, all VR applications have a modular structure. Let us show that IMHAP
platform (Liang et al, 2007) is divided into three components: Model, Viewer, and Controller.
Of course, the controller implements the interactions between user and the model. The
architecture of a remote mechatronic virtual laboratory (useful for virtual training in
robotics) is shown in Figure 1, based on the presentation from (Pasc et al, 2007) where
the interaction is assured by commands and signals. Such architecture is suitable also for
medical surgery at distance, or operation at distance, according to (Albeanu, 2007).
Another kind of architecture is the one proposed by (Boulic et al, 2003) which
consists in layers. For instance, the H-Anim architecture contains an walking engine
having three layers: generic walk pattern, gait personification, and walking trajectory
controllers. All of them are designed in order to maintain the coherence of the model.
The applications designed in order to model some body parts for virtual bodies use
different ideas based on: volume imaging technologies, surface rendering and hybrids
models, and volume rendering, as (Waterworth, 1999) reviewed.
Other architectures are behaviour oriented like those proposed by (Popovici, 2005),
or component oriented like for VR based applications developed by (Haller, 2001).
Anyway, our biographical research establish a large collection of contributions (not all
mentioned in references), but the basic ideas are those already presented above.


3. Simulation techniques for VR applications

Simulation is the second stage of every VR application, the first stage being the model
development. The simulation techniques depend on mathematical models associated to
the virtual model. The simulation models used for implementing the behaviour of different
real/virtual actors/systems are based on discrete or continuous mathematical model.
When considering the simulation technology, the VR project manager will consider
both virtual and physical systems architectures and their integration. For some VR applications,
like those of collaborative nature, are necessary distributed simulation methods. The final
stage deals with the validation of the simulation model and comparison of different
simulation areas (such as vehicle, weather, medical, industrial, and entertainment).
Various mathematical methods are required in different simulation scenarios
(matrix transformations, algebra, trig, complex numbers), as well as open-loop and
closed-loop system theory, discrete versus continuous simulation, the use of databases in
simulations, and the necessary real-world physics/biology/chemistry etc. The references
(Bell and Fogler, 1997), (Dimitropulos et al, 2008), (Jung et al, 2005), (Metze et al,
2005), (Souza et al, 2007) and (Thelen and Anderson, 2006) are only some of a huge
scientific literature dedicated to different aspects on simulation for different VR applications.
In the following we establish the main steps of a any simulation scenario: (1)
establish the unit of time or/and distance depending on application; (2) establish the
simulation time (how long?); (3) simulation start clock and uniform/variable time-step
length; (4) setting the objects behaviour (movement, collision avoidance, …), and (5)
generation, analysis and storage of new information.
University of Bucharest and Ovidius University of Constanta


66
Some VR applications use backward simulation. The backward simulation asks to
start from a current or final state and to move backwards in time to an initial state in order
to determine the sequence of actions (the trajectory, the path) for moving the system from
the initial state to the current final state. Other VR applications use forward simulation
based approaches that usually execute a single forward pass through time based on some
dispatching rules to develop the sequence.
These techniques can be used also for motion generation required by some VR
applications by automated or interactive control. Both keyframing and procedural
methods can benefit from backward/forward simulations. Keyframing asks for key
positions of objects to be animated and then interpolation is necessary to identify the
positions in-between frames. Inverse kinematics (Zhao & Badler, 1994; Badler et al,
1999; etc) and different kind of interpolation procedures (Albeanu, 1999; Badler et al,
1999; Magnenat-Thalmann & Thalmann, 2004); etc) are implemented during simulation
and motion generation. For particle systems, procedural methods can be developed based
on laws of physics to generate motion. Not only individual objects but also groups of
objects can be moved together during motion generation. Capturing or graphical design
can also be used as INMEDEA uses in its web-based medical simulator.
If considering human walking, the simulator divides the simulation time in
succession of phases: right takeoff (RT), right footstrike (RF), left takeoff (LT), left
footstrike (LF), and its implementation will ensure that the body parts motion and contact
with terrain or stairs looks realistic, and will solve pushing/collisions and other kind of
interventions (Figure 2). (Thelen & Anderson, 2006) developed a powerful methodology
based on forward dynamic musculoskeletal simulation model. According to (Multon et al,
1999) some methods will be mixed depending on the VR application type.



Figura 2. Time rule during RT-RF-LT-LF cicle

Of course, simulations models used for VR applications in chemistry (Bell & Fogler,
1997) or environment protection (Albeanu, 2007) are completely different from human
walking simulation, but the main ideas about controlling the sequence of events remain.


4. Some mathematical models

Mathematical modelling is important not only for scientific aspects in industrial
and business applications, but also for virtual learning and entertainment. Not only
geometric transformations (2D, 3D), viewing transformations (parallel or
perspective projection), clipping and hidden line or surface removal, well known
in computer graphics, but mathematical models dependent on the concrete
applications are required in order to create realistic behaviour of the objects belonging to
a simulated environment.
In the following some models concerning trajectory generation, surface-terrain
generation and object generation will be detailed.
RT-RF LT-LF

The 3
rd
International Conference on Virtual Learning, ICVL 2008

67
4.1. Trajectory/terrain modelling

For the modeling of a walking/evolutionary trajectory, curves interpolated from
data, or approximation curves can be used. Special curves could be obtained using
trigonometric interpolation as described in (Albeanu, 1999). However, interproximation
can be used as a mixed interpolation-approximation methodology as described in
(Falcidieno & Kunii, 1993) by Cheng and Barsky (pag. 359). This methodology can also
be used to model closed shapes.
Let D = {D
i
| i = 1, 2, …, n} be a set of 2D data (points described as D
i[j]
= P
j
(x
j
, y
j
), or
rectangles described as D
i[j]

= [a
j
, b
j
]x[c
j
, d
j
], j = 1, 2, …, n. A common interproximation
scheme uses cubic B-splines to fit the data set D.
The cubic B-spline curve is a piecewise curve of n+m-1 segments, requiring n+m+6
interpolating knots denoted by T = {t
-2
, t
-1
, t
0
, t
1
, …, t
n+m+3
}, where t
-2
= t
-1
= t
0
= t
1
= 0,
t
n+m
= t
n+m+1
= t
n+m+2
= t
n+m+3
= 1, and t
i
= t
i-1
+ d
i
, where d
i
is given according to the
centripetal model of Lee (cited in (Falcidieno & Kunii, 1993)):

1 - m n i 2 ,
2
2 / 1
1
2 / 1
1
+ ≤ ≤


=

+
=


n m
j
j j
i i
i
A A
A A
d ,

with A
i
being D
i[j]
(the point or the centre of the rectangle), j = 1, 2, …, n.
The cubic B-spline defined on [0, 1] can be represented as

, ) ( ) (
1
2
3 , 2 ∑
− −
− =

=
n m
i
i i
u N C u S [0,1] u Î ,

where C
i
are control points and N
i,3
(.) are normalised cubic B-splines defined related to
the sequence of knots T.
Other trajectories can be obtained using trigonometric piecewise functions. If F
i
(.), i
= 1, 2, 3, 4 are Hermite trigonometric polynomials having parameters α and β, the
trigonometric curve with endpoints P
α
and P
β
, and derivatives P’
α
and P’
β
, has the
representation: h
α
,
β
(t) = P
α
F
1
(t) + P
β
F
2
(t) + P’
α
F
3
(t) + P’
β
F
4
(t), t ∈[α, β].
A suitable representation of terrain regions in order to apply collision detection, or
finding contact points with some objects is based on Bézier rectangles.





Figure 3. Terrain generation by Bézier rectangles with texturing
University of Bucharest and Ovidius University of Constanta


68
A Bézier rectangle of degree mxn has the form

∑∑
= =
≤ ≤ =
m
i
n
j
j i n j m i
P v B u B v u S
0 0
, , ,
1. v u, 0 , ) ( ) ( ) , (

where B
i,m
(.), respectiv B
j,n
(.) are univariate Bernstein polynomials of degree m, respectiv,
n, and P
i,j
are the control points of the Bézier rectangle.
For some applications a texture is applied in order to obtain a realistic view (Figure
3), but from many algorithmic tasks, only the skeletal of the terrain is required.


4.2. Object modelling

Solid physical objects can be represented as a combination (union, intersection,
difference) of primitives like cubes, spheres, cones, cylinders, tetrahedrons or quadratic
pyramids, etc. according to the Constructive Solid Geometry (CSG) based methodology.
Other methods are: spatial enumeration, cell decomposition, boundary representation, and
primitive instancing. Some useful models consider super-primitives like super-elipsoid
and super-toroid objects and other entities obtained by mathematical transformations
(translation along, rotations, …, etc) or sweep methods.
The most natural way to represent a CSG model is the CSG tree:

<CSG tree> ::= <primitive> |
<CSG tree> <set operation> <CSG tree> |
<CSG tree> <rigid motion>


where <primitive> is an instance of one of the primitives of the primitive data
base, <rigid motion> is either a translation or a rotation, and <set operation> is either
∪, ∩, or ÷ (setminus).
To animate the scene, both a static description and information about object
movement is required. The information about the movement is described in an articulated
model (the objects are connected by joints in a hierarchical structure). For some objects
the motion is determined by rules (for instance, the laws of physics), specified also for
deformable entities.
A special case for VR applications deals with trivariate data modelling, that means
the construction of a function F(x, y, z) which interaproximates the relationship implied
by the data (x
i
, y
i
, z
i
; F
i
). If (x
i
, y
i
, z
i
) are interior points of some object, the model will
provide information about attributes of other points belonging to the object.
This information is useful for a realistic rendering of the animated scene. A
common method to identify a model uses the distance function approach (least square).
Piecewise Hermite form of the spline models can also be used, and the coefficients will
be identified by solving the obtained linear system of equation.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

69
5. Conclusions

This paper described the main principles of the simulation models used to
implement virtual reality applications. There are presented the architecture of VR
systems, VR applications in different fields, including medicine, an introduction to
simulation techniques and a set of mathematical models for creating virtual scenes.
The complexity of the subject is large and a strong mathematical background is
necessary. Also, implementing VR applications asks for recent information technologies
resources including virtual reality hardware and software tools.



REFERENCES


ALBEANU, G. (1999), On the Geometric Modeling of Curves by Trigonometric Polynomials, Annals of
Bucharest University, Mathematics-Informatics Series 48, 1, 55-60.
ALBEANU, G. (2007), Lecture on Simulation Models for VR Applications, VRRM-2007: International
Workshop on Virtual Reality in Rehabilitation Medicine, University of Medicine and Pharmacy “Carol
Davila”, Bucharest, Romania, September 24-25, 2007, Manuscript.
ALBEANU, G. (2006), Modelling and Programming VR Applications – an ICT undergraduate course,
Proceedings of ICVL 2006, Bucharest, 83-90, http://fmi.unibuc.ro/cniv/2006/disc/icvl/
documente/pdf/met/2_albeanu.pdf.
ANDERSON, F. C. and PANDY, M. G. (2001), Dynamic Optimization of Human Walking. Journal of
Biomechanical Engineering 123, 381-390.
ANDERSON, F. C., ARNOLD, A. S., PANDY, M. G., GOLDBERG, S. R. and DELP S. L. (2006),
Simulation of Walking. In Human Walking, Williams and Wilkins, 3rd Edition, Baltimore,
http://nmbl.stanford.edu/publications/pdf/Anderson2006.pdf.
BADLER, N. A., PHILIPS, C. B. and WEBBER B. L. (1999), Simulating Humans: Computer Graphics,
Animation, and Control. Oxford University Press, New York.
BELL, J. T. and FOGLER, H. S. (1997), The application of virtual reality to chemical engineering education,
Proceedings of the 1997 International Conference on Simulation in Engineering Education (ICSEE
‘97), January 12-15, 1997, Simulation Series 29(2), Society for Computer Simulation, San Diego.
BOULIC, R., GLARDON, P. and THALMANN, D. (2003), From measurements to model: the walk engine,
Proc. of 6th Conf on Optical 3D Measurement Techniques. The 6th Conference on Optical 3D
measurement Techniques, http://infoscience.epfl.ch/getfile.py?docid=13105&name=Boulic
_Glardon_Thalmann_CO3DMT_03&format=pdf&version=1.
DIMITROPOULOS, K., MANITSARIS, A. and MAVRIDIS, I. (2008), Building Virtual Reality
Environments for Distance Education on the Web: A Case Study in Medical Education. International
Journal of Social Sciences 2, 1, 62-70.
FALCIDIENO, B. and KUNII, T. L. (eds.) (1993), Modeling in computer Graphics, Springer Verlag.
FUCHS, P. and MOREAU, G. (2003), Le traité de la réalité virtuelle, Vol. 2: Création des environments
virtuels & applications, Ecole des mines de Paris, Paris.
HAGGEN, H., MÜLLER, H. and NIELSON, G. M. (eds.) (1993), Focus on Scientific Visualization,
Springer Verlag.
HALLER, M. (2001), A component oriented design for a VR based application. International Workshop on
Structured Design of Virtual Environments and 3D-Components at the Web3D 2001 Conference,
Paderborn, Germany, February 19th, 2001.
University of Bucharest and Ovidius University of Constanta


70
Inmedea simulator: http://www.inmedea-simulator.net, INMEDEA GmbH, Tübingen.
JUNG, B, AHAD, A. and WEBER, M. (2005), The Affective Virtual Patient: An e-Learning Tool for social
Interaction Training within the Medical Field, Training Education & Education International Conference
(TESI 2005). Proceedings Nexus Media, March, http://isnm.de/~aahad/Downloads/AVP_TESI.pdf.
LIANG, C.-H., TAO P.-C. and LI, T. Y. (2007), IMHAP-An experimental Platform for Humanoid Procedural
Animation, Proceedings of the third International Conference on Intelligent Information Hiding and
Multimedia Signal Processing, Tainan, Taiwan, http://imlab.cs.nccu.edu.tw/paper/2007.11iihmsp2007.pdf.
MAC NAMEE, B., DOBBYN, S., CUNNINGHAM, P. and O’SULLIVAN C. (2003), Simulating Virtual
Humans Across Diverse Situations. Proceedings of Intelligent Virtual Agents ’03, 159-163,
http://www.comp.dit.ie/bmacnamee/papers/SimulatingVirtual HumansAcrossDiverseSituations.pdf.
MAGNENAT-THALMANN, N. and THALMANN, D. (editors) (2004), Handbook of Virtual
Humans, John Wiley & Sons, Ltd.
METZE, J., NEIDHOLD, B. and WACKER, M. (2005), Towards a general concept for distributed
visualisation of simulations in Virtual Reality environments, IPT & EGVE Workshop, 1–12.
MULTON, F., FRANCE, L., CANI-GASCUEL M.-P. and DEBUNNE, G. (1999), Computer Animation
of Human Walking: a Survey. The Journal of Visualization and Computer Animation 10, 39-54
NG-THOW-HING, V. (2001), Anatomically-based models for physical and geometric
reconstruction of humans and other animals, PhD Thesis, University of Toronto.
PASC, I. M., TARCA, R. C., POPENTIU-VLADICESCU, FL. and ALBEANU, G. (2007), On
Designing Virtual Environments Based on Intelligent Mechatronic Systems, ISSAT
International Conference on Modeling of Complex Systems and Enviroments, 16-18 July
2007, Ho Chi Minch City, Vietnam, 126-130.
POPOVICI, D. M. (2005), Behavioral aspects and behavior-oriented architectures in 3D virtual
environments, An. St. Univ. Ovidius Constanta 13, 2, 61-74.
SILVERT, W. (2000), Modelling as a discipline. Int. J. General Systems 30, 261-282.
SOUZA, D. F. L., CUNHA, I. L. L., SOUZA, L. C., MORAES, R. M. and MACHADO, L. S. (2007),
Development of a VR Simulator for Medical Training Using Free Tools: A Case Study.
Proc. of Symposium on Virtual and Augmented Reality (SVR'2007), 100-105.
THELEN, D. G. and ANDERSON, F. C. (2006), Using computed muscle control to generate
forward dynamic simulations of human walking from experimental data. Journal of
Biomechanics 39, 1107 – 1115.
ARCĂ, R. C., IVĂNESCU, A., BARDA, I., ALBEANU, G., PASC, I. and POPENIU VLĂDICESCU FL.
(2008): Augmented reality used for a remote robot control, The 12th World Multi-
Conference on Systemics, Cybernetics and Informatics (WMSCI 2008), Orlando, Florida
(Jun 29
th
-July 2
nd
), in press.
WATERWORTH, J. A. (1999), Virtual Reality in Medicine: a survey of the state of the art,
http://www.informatik.umu.se/~jwworth/medpage.html.
ZHAO, J. and BADLER, N. I. (1994), Inverse Kinematics Positioning Using Nonlinear Programming
for Highly Articulated Figures. ACM Transactions on Graphics 13, 4, 313-336.



The 3
rd
International Conference on Virtual Learning, ICVL 2008

71
Modeling of Errors Realized by a Human Learner
in Virtual Environment for Training

Thanh Hai Trinh
1, 2
, Cédric Buche
1
,
Jacques Tisseau
1


Université Européenne de Bretagne – ENIB – LISyc – CERV
Technopôle Brest-Iroise, 29238 Brest Cedex 3, FRANCE
(2) Institut de la Francophonie pour l’Informatique
42 Ta Quang Buu, Ha Noi, VIETNAM
E-mail: tthai@ifi.edu.vn; buche@enib.fr; tisseau@enib.fr


Abstract
This study focuses on the notion of erroneous actions realized by human learners in
Virtual Environments for Training. Our principal objective is to develop an
Intelligent Tutoring System (ITS) suggesting pedagogical assistances to the teacher.
For that, the ITS must obviously detect and classify erroneous actions produced by
learners during their realization of procedural and collaborative work. Further, in
order to better support human teacher and facilitate his comprehension, it is
necessary to show the teacher why learner made an error. Addressing this issue, we
firstly model the Cognitive Reliability and Error Analysis Method (CREAM). Then,
we integrate the retrospective analysis mechanism of CREAM into our existing ITS,
thus enable the system to indicate the path of probable cause-effect explaining
reasons why errors have occurred.

Keywords: Intelligent tutoring system, Erroneous actions, Retrospective analysis.


1. Introduction

In order to simulate procedural and collaborative work, we previously developed
the model MASCARET (Multi-Agent System for Collaborative Adaptive and Realistic
Environment for Training) where human learners and agents collaborate to realize a mission
(Querrec et al., 2004). Learners are gathered in team consisting of several predefined
roles, every role contains a number of tasks to be realized by learners with accurate
resources. During realisation of the tasks, it is essential to take into account that human
learners could make erroneous actions in comparing to their predefined correct procedure.
In (Buche and Querrec, 2005), we have proposed a model of Intelligent Tutoring
System (ITS) whose principal objective is to suggest pedagogical assistances to the
teacher adapted to the simulation context and to the learner’s behaviours (including
erroneous actions). However, this works exclusively concerns errors detection and
tagging. Once erroneous actions are detected in our existing ITS, it were be classified in
different types (see Figure 1a) whose explications are based on a knowledge base on
University of Bucharest and Ovidius University of Constanta


72
classical errors. In order to better support the teacher and facilitate his comprehension, it
lacks a model that could explain reasons why the learner made an error.
Our approach bases on the Cognitive Reliability and Error Analysis Method
(CREAM) in Human Reliability Analysis field (Hollnagel, 1998). This approach
proposed a classification scheme which makes a distinction between observations of
errors (phenotypes, see Figure 1b) and its causes (genotypes) classified in three
categories: M(an), T(echnology) and O(rganization). The causal links between
phenotype-genotype are represented using a number of consequent-antecedent links.
Finally, the scheme could be associated with both a method of retrospective analysis (the
search for causes) and a performance prediction method. However, in our goal of
erroneous actions detection and then searching for the causes, we interested in human
learner’s performance analyses, in other words, in retrospective analyses.

TeamError
Error
TeamProceduralError
ProceduralError ActionError UsageError

Erroneous
action
Timing Duration
Sequence
Object
Force Direction
Distance
Speed
Too early, too
late, omission
Too long,
too short
Reversal, repetition,
commission,
intrusion
Wrong action,
wrong object
Too much,
too little Wrong direction
Too far,
too short
Too fast,
too slow

Figure 4a. Errors types in ITS
(Buche and Querrec, 2005)
Figure 1b. Dimensions of error modes
(Hollnagel, 1998)

Implementation of CREAM was object in the work of (El-Kechaï, 2006, 2007)
which firstly proposed a task model named METISSE in order to recognize learner’s
plans in Virtual Environments for Training (VET), then this model could be used to
detect for erroneous actions according to classification of Hollnagel. Nevertheless,
implementation of METISSE was not complete, and integration of CREAM into a really
ITS was not performed.
In this paper, we will firstly propose an approach to model CREAM (section 2). Next,
in section 3, we will present the integration of retrospective analysis mechanism of CREAM
into our existing ITS as well as our evaluation. Finally, section 4 summarizes the present work.


2. Implementation of CREAM

2.1. Classification Scheme Representation

There are several graphic tools that permit to keep track of analyses processes such
as CREAM Navigator developed by (Serwy and Rantanen, 2007). However, this
navigator is completely closed in the sense that it does not maintain an explicit
The 3
rd
International Conference on Virtual Learning, ICVL 2008

73
representation of possible errors modes and probable causes. For that, (El-Kechaï and
Després, 2007) proposed using a rules base for represent consequent-antecedent links,
hence the search for the causes was executed by backward inferences. Limitation of this
method obviously lies on the performance of inference mechanism, other problem maybe
occurs in adding, removing another potential errors that will demand a considerable
modification on the rules base. For our development, as suggested in (Hollnagel, 1998),
we intent to separate the analysis method (cf. section 2.3 and 2.4) and the representation
of errors modes using a group of four data files in format XML detailed below:
– Questionnaire.xml: proposing to represent a list of questions from which we
could evaluate the Common Performance Conditions (see section 2.2 in following)
– Phenotype.xml: proposing to maintain the phenotypes and its antecedents

<Phenotypes>
<Phenotype name="Time/During" description="...”>
<GeneralAntecedents>
<item>Inadequate plan</item>
<item>Inattention</item>
</GeneralAntecedents>
<SpecificAntecedents>
<item>Earlier omission</item>
</SpecificAntecedents>
</Phenotype>…
</Phenotypes>

– Genotype.xml: containing all possible causes classified in three groups (M,T,O), each
group is then detailed into several categories. The important point is that this data file also
represents relations between each consequent and its antecedents

<Genotypes>
<Group name="Man">
<Category name="planning">
<GeneralConsequent name=" Inadequate plan" description="...”>
<GeneralAntecedents>
<item>Distraction</item>
<item>Excessive demand</item>
</GeneralAntecedents>
<SpecificAntecedents>
<item>Error in goal</item>
<item>Inadequate training</item>
</SpecificAntecedents>
</GeneralConsequent>
</ Category >
</ Group >…
</Genotypes>

University of Bucharest and Ovidius University of Constanta


74
– Repartition.xml: proposing to determine repartition of specific antecedents in
three factors (M,T,O) which serves to initialize the mass of each specific antecedent as a
probable cause (cf. section 2.4)

< Repartition >
<item name=" Earlier omission” group="Man" description="…"/>
<item name=" Message misunderstood” group="Technology" description="…"/>…
</ Repartition >

Finally, in considering that CREAM is naturally a flexible method and adaptable to
different analysis contexts, this strategy of classification scheme representation permits
customize the scheme without any modification on analysis method.


2.2. Define the Common Performance Conditions (CPC’s)

In CREAM, Hollnagel highlighted that the context strongly influence human
actions. It is therefore essential to take into account the description of virtual environment
in which the human learner is immersed. The objective is to determine how each factor
(M,T,O) influences the training context. Here, we are inspired from the proposition
presented in (El-Kechaï and Després, 2007) using a predefined questionnaire which will
be answered by the teacher before training session:

<Questionnaire>
<Question name="The visual quality of the interface is it bad?" group="Technology" answer="No"/>
<Question name="Does the learner have concentration trouble?" group="Man" answer="No"/>…
</Questionnaire>

Next, each factor will be assigned one coefficient calculated using formula below:

[1]
answers of number Total
group to associated answers of Number
i) (
Yes
i Yes
group t Coefficien =


where group i is respectively in (Man, Technology, Organization). These values permit
define the most probable factor leading to erroneous actions.


2.3. Modelling of Consequent-Antecedent Relations

One advantage of CREAM lies on its recursive analysis approach, rather than
strictly sequential in compare with other traditional analysis methods. So that, it also
conducts to a non-hierarchical data structure to connect the direct as well as indirect links:
The 3
rd
International Conference on Virtual Learning, ICVL 2008

75
(i) between a phenotype and its antecedent; and (ii) between a consequent and its antecedents.
Figure 2 shows our model to represent the connection between consequent – antecedent.

+addAntecedent()
+addConsequent()
+calculateMass()
-_name : string
-_group : string
-_category : string
-_description : string
-_mass : double
-_terminal : bool
-_list _antecedent
-_list _consequent
Node
+getQuestionnaire()
+getPhenotypes()
+getSpecificAntecedents()
+getGeneralConsequents()
+getGeneralAntecedents()
Util
1
0..1
0..1
0..*
1
1
«uses»
Questionnaire.xml
Phenotype.xml
Genotype.xml
Repartition.xml
+getAntecedentFromPhenotype()
+getGenotypeFromAntecedent()
+findSpecificAntecedentRepartition()
+createGraphFromPhenotype()
+findListTerminal()
+sortListTerminal()
-_graph : string
GenotypeAnalyzer

Figure 2. Our UML diagram for modeling consequent-antecedent links

Here, we are going to construct a causal graph where we use the term “node” to
point to either a consequent or an antecedent. Each node is described by its name; the
group of errors modes that it is associated and its category in group; the description in
text helps better explain the error’s semantics in particular context. The boolean attribute
terminal permit to identify if that is a terminal-cause or not. The most important is that,
each node contains two lists: one includes its antecedents, other points to its consequents,
in others words, they represent edges in/out one node in the causal graph. At last, each
node must also include a value of mass which represent the certitude of choosing this
node as a probable cause. The two methods addAntecedent() and addConsequent() serve
for maintaining the two lists of antecedents and consequents of one node. Note that once
a node calls the method addAntecedent() serving for adding a “parent” node like one of
its antecedents, this node will also add itself to the consequents list of the “parent” node
(using the method addConsequent() of the parent node) , the value of the attribute
terminal then will be set to false.


2.4. Search for the Causes

Finally, the retrospective analysis is executed by a GenotypeAnalyzer containing
graph attribute which is initialized by pointing to the initiating phenotype (“root” node),
then the analyzer calls accurate methods to find the “root” causes (the nodes with the
attribute terminal having value false). This mechanism is presented below:

Input: Phenotype of erroneous action
Initialization: Construct the “root” node pointing to phenotype input
University of Bucharest and Ovidius University of Constanta


76
Step 1: Read from file Phenotype.xml, find all general antecedents of phenotype input
For each antecedent Do
Add it into antecedents list of “root” node
Step 2: For each unvisited node in the graph Do
– Find its antecedents from file Genotype.xml & add them to list
– Return step 2. This recursive search terminates when the nodes
selected is a specific antecedent node or a general consequent node
without antecedents.

With this algorithm, we finally attain a causal network where each node is
associated with its antecedents and consequents. The “leaves” are terminal nodes (or
“root” causes) whose antecedents list is empty. In order to calculate the certitude of
choosing each node as a probable cause, we inherit the proposition presented in (El-
Kechaï and Després, 2007) using Dempster-Shafer’s evidence theory:

[2]
( )

∑ ∈ ∀
∈ ∀
|
|
|
¹
|

\
|
=
) (
} , , {
x ) (
) (
x )) ( ( ) (
a Cons c
O T M i
ic
n i t coefficien
c mass
a g t coefficien a mass

where:

- mass(a): mass of antecedent a
- g(a): group of a
– coefficient (i): coefficient of group i calculated in formula [1]
- Cons(a): consequents list of a – n
ic
: number of antecedents of c classified in group i


3.Integration of Retrospective Analysis into our existing ITS

3.1. Learner’s Plans Recognition

In order to detect the erroneous actions realized by a human learner, it is
indispensable to know: (1) the learner’s activities in the past; (2) his current action (in the
meaning that the action has just been done); and (3) the actions that the human learner
intents to do in according to a predefined correct procedure. Our existing ITS as proposed
in (Buche and Querrec, 2005) bases on the model MASCARET (Querrec et al., 2004)
where we used an approach using multi-agent system to simulate collaboration between
human learners and agents during their realization of missions. Learners are gathered in
team consisting of several predefined roles, every role contains a number of tasks
associated eventually with accurate resources, every leaner also owns an epistemic
memory containing all actions realized in the past, etc. Finally, we could retrieve from
MASCARET following informations relating to learner’s plan in VET:
The 3
rd
International Conference on Virtual Learning, ICVL 2008

77
– action(s) before: learner’s action(s) in the past (note that, in MASCARET, every
action is eventually associated with its accurate resource(s))
– current action: action has just been done by learner
– action(s) correct (according to role): action(s) must be done by learner in his role(s)
– action(s) correct (according to plan): action(s) may be done by learners in the
context. Here, it is essential to make distinction betweens “action(s) correct according to
role” and “action(s) correct according to plan”. In the first case, because the learner
could play several roles, it represents all correct actions that the system expects from the
learners. The second one concerns the cases where there are more than one learner in
VET to realize together a mission. Therefore, in this case, it is possible that a leaner
performs a correct action according to the plan but it is not correct in compare to his role.
– next correct action(s) in the role: next action(s) must be done by learner in his role(s)
– full correct plan: description of all accurate actions (associated with resources) in
predetermined procedure that the learner must respect.
In next section, we present our mechanism for mapping erroneous actions detected
by our existing ITS with Hollnagel’s classification scheme of errors modes.


3.2. Classification of Erroneous Actions according to the Scheme of CREAM
Erroneous Actions in Phenotype “Sequence”

According to Hollnagel, performing an action at the wrong place in a sequence or
procedure is a common erroneous action, and it is more realistic in our context of
simulation of procedural and collaborative work. The “Sequence” problem consists of
several specific effects: Omission (an action was not carried out); Jump forward/ Jump
backwards (actions in a sequence were skipped/carried out again); Repetition (the
previous action is repeated); Reversal (the order of two neighbouring action is reversed);
Wrong action (an extraneous or irrelevant action is carried out). We present in following
our mechanism to detect erroneous actions in phenotype “Sequence”:

– If current action exists in action(s) correct according to role:
this is a correct action (phenotype Sequence does not occur).
Else: + If current action does not exist in action(s) correct (according to plan):
specific effect = “Wrong action”
Else: * If current action exist in last action before: specific effect = “Repetition”
* Compare the relative order of current action to the order of next correct action(s) in
the role using the full correct plan:
– If id_current_action < id_correct_action_in_role:
specific effect = “Jump backwards and/or Omission ”
Else: specific effect = “Jump forward and/or Omission”
– If id_current_action = id_correct_action_in_role +1:
specific effect = “Reversal ”
University of Bucharest and Ovidius University of Constanta


78
Erroneous Actions in Phenotype “Wrong object”

In (Hollnagel, 1998), the author clarified that “action at wrong object” is one of the
more frequent error modes, such as pressing the wrong button, looking at the wrong
indicator, etc. In our context, during realisation of collaborative work, it is possible that
learner performs a correct action but on a wrong object. Therefore, the detection of
erroneous actions in phenotype “Wrong object” must be implemented independently with
the detection of phenotype “Sequence”. This phenotype is detailed into following specific
effects: Neighbour/Similar object (an object that is proximity/similar to the object that
should have been used); Unrelated object (an object that was used by mistake).
In order to detect erroneous actions in phenotype “Wrong object”, we use the same
principle presented in the case of phenotype “Sequence” by using following informations
retrieved from model MASCARET:
– current resource: resource associated with current action
– resource(s) correct (according to role): resource(s) must be used by learner in his role(s)
– resource(s) correct (according to plan): list of resource(s) associated with all
action(s) in action(s) correct according to plan.
Our algorithm is detailed in following:

– If current resource exists in resource(s) correct according to role:
this is a correct resource (phenotype Wrong object does not occur).
Else: + If current resource does not exist in resource(s) correct (according to plan):
specific effect = “Unrelated object”
Else specific effect = “Neighbour and/or Similar object ”


Erroneous Actions in Phenotype “Time/During”

The phenotype “Time/During” is divided in several specific effects: Too early/ Too
late (an action started too early/too late); Omission (an action that was not done at all);
Too long/Too short (an action that continued/was stopped beyond the point when it
should have been). Hollnagel noted that the error modes of timing and duration refer to a
single action, rather than to the temporal relation between two or more actions. In our
context, the realization of tasks in model MASCARET is sequential, therefore, an action
is considered to be too early when it was realized before several actions in plan; also,
action(s) are considered to be omitted when they were not carried out.
Finally, in order to detect erroneous actions in phenotype “Time/Durring”, we
propose that:
– action having specific effect ““Jump forward” also has specific effect “Too early”
– action described by specific effect “Omission”(in error mode “Sequence”) will be
considered as an action having specific effect “Omission” (in error mode “Time/During”)
The 3
rd
International Conference on Virtual Learning, ICVL 2008

79
3.3. Experiment & Results

In order to evaluate our integration of retrospective analysis into ITS, we take place
in GASPAR application (Marion et al., 2007) whose objective aims at simulate aviation
activities by virtual reality. We use the classification scheme of error modes proposed in
(El-Kechaï and Després, 2007) which were particularly adapted to VET. Table 1
illustrates results of retrospective analysis for the phenotype Sequence.

Table 1
Causal links of phenotype "Sequence"

Coefficient (M,T,O) Causal links
1, Design failure (0.125) -> Inadequate scenario (0.125) -> Sequence
2, Adverse ambient condition (0.125) -> Inattention (0.125) -> Sequence 0.333 - 0.333 - 0.333
3, Long time since learning (0.042) -> Memory failure (0.125) -> Sequence
1, Other priority (0.2) -> Memory failure (0.2) -> Sequence
2, Error in mental model (0.067) -> Faulty diagnosis (0.2) -> Sequence 1 - 0 - 0
3, Erroneous analogy (0.067) -> Faulty diagnosis (0.2) -> Sequence
1, Equipment failure (0.1) -> Access problems (0.5) -> Sequence
2, Distance (0.1) -> Access problems (0.5) -> Sequence 0 - 1 - 0
3, Localisation problem (0.1) -> Access problems (0.5) -> Sequence
0 - 0 - 1 1, Noise (1) -> Communication failure (1) -> Sequence

We change coefficients of three factors (M,T,O) for evaluating how CPC’s
influence the analysis result. For each phase in analysis process, we select and display the
most probable cause by ordering mass values.


4. Conclusion and Future Work

In this paper, we proposed an approach to modelling the Cognitive Reliability and
Error Analysis Method (CREAM). We separated the representation of classification
scheme of erroneous actions and the analysis method; therefore, our description of errors
modes is adaptable to different training context without any modification on analysis
method. We started by defining the Common Performance Conditions, then the direct and
indirect relations between consequent-antecedent are modelled using a non-hierarchical
data structure. Finally, the most probable cause-effect links could be found using
Dempster-Shafer’s theory presented in (El-Kechaï and Després, 2007).
In order to integrate the retrospective analysis described above into our existing
ITS, we based on the model MASCARET to retrieve information concerning learner’s
plans and then detect erroneous actions. Finally, we presented our proposition to mapping
erroneous actions with Hollnagel’s classification. The experimental results in GASPAR
University of Bucharest and Ovidius University of Constanta


80
project are also presented. So that, in addition to the detection and tagging of erroneous
actions, the ITS could furthermore indicate the path of probable cause-effect explaining
reasons that the errors occur.
In the future work, we will concentrate our attention on evaluation of MASCARET
so that this model could permit to describe more complex tasks in taking into account
other factors such as force, distance, speed, direction, etc. Hence, other different types of
errors modes could be detected and then explained using the retrospective analysis.



REFERENCES


QUERREC R., BUCHE C., MAFFRE E., and CHEVAILLIER P. (2004), Multiagents systems for
virtual environment for training: application to fire-fighting. International Journal of
Computers and Applications (IJCA), June 2004, 25-34.
BUCHE C. and QUERREC R. (2005), Intelligent tutoring system for MASCARET, in Simon
Richir and Bernard Taravel, editors, 7th Virtual Reality International Conference
(VRIC'05), April 2005, Laval, France, 105-108.
HOLLNAGEL, E. (1998), Cognitive Reliability and Error Analysis Method, Oxford: Elsevier
Science Ltd.
EL-KECHAÏ N. and DESPRÉS C. (2007), Proposing the underlying causes that lead to the
trainee's erroneous actions to the trainer, in EC-TEL: European Conference on Technology
Enhanced Learning, September 2007, Crète (Grèce), 41-55.
EL-KECHAÏ N. and DESPRÉS C. (2006), A Plan Recognition Process, Based on a Task Model,
for Detecting Learner's Erroneous Actions, in Intelligent Tutoring Systems ITS 2006, June
2006, Jhongli (Taïwan), 329-338.
MARION N., SEPTSEAULT C., BOUDINOT A. and QUERREC R. (2007), GASPAR: Aviation
management on an aircraft carrier using virtual reality, in Cyberworlds 2007.
SERWY R. D. and RANTANEN E. M. (2007), CREAM Navigator http://www.ews.uiuc.edu/~serwy/
cream/v0.6beta/, [version 0.6, September, 2007]


The 3
rd
International Conference on Virtual Learning, ICVL 2008

81
Architecture and Working Principles of the Concept Map Based
Knowledge Assessment System

Marks Vilkelis
1
, Alla Anohina
1
, Romans Lukashenko
1


(1) Department of Systems Theory and Design, Riga Technical University
1, Kalku Str., Riga, LV-1658, LATVIA
E-mail: {markvilkel@inbox.lv, alla.anohina@rtu.lv, LREXPress@inbox.lv}


Abstract
The paper describes the concept map based knowledge assessment demonstrating its
main functionality on the basis of screenshots and presenting the three-tier client-
server architecture of the system in terms of components, their functions and
interaction. Underlying conceptions and current development directions related to
the implementation of learner’s supports are discussed as well.

Keywords: Concept map, Assessment system, Learner’s support


1. Introduction

Rapid development of information and communication technologies has led to the
appearance of a new generation of young people who cannot imagine their life without
the use of computers. The computer serves not only as an instrument of acquisition of
necessary information or as an environment for communication and entertainment, but
also as a tool for learning. This is a reason why the great part of educational institutions
all over the world introduce different information and communication technologies, such
as e-learning environments, videoconferences, intelligent tutoring systems, etc., in the
process of teaching and learning.
The Department of Systems Theory and Design of the Faculty of Computer Science
and Information Technology of Riga Technical University has been developing the
concept map based knowledge assessment system since the year 2005. The system has
twofold goals in the context of the integration of technology into the traditional
educational process: 1) to promote learners' knowledge self-assessment, and 2) to support
the teacher in the improvement of the learning course through systematic assessment of
learners' knowledge and analysis of its results. The goals are reached by the use of
concept maps as an assessment tool. At the moment the system has reached the certain
level of maturity concerning its architecture and working principles which are presented
in this paper.
The paper is organized as follows. Section 2 gives an overview of the system. The
architecture of the system presenting its main components and technologies is described
in Section 3. Section 4 demonstrates an example of the system’s operation by means of
University of Bucharest and Ovidius University of Constanta


82
screenshots. The current development directions related to the implementation of
learner’s support are discussed in Section 5. The paper ends with Conclusions.


2. Overview of the System

As it was mentioned in Introduction concept maps are used as an assessment tool in
the developed system. According to (Cañas, 2003) they can foster the learning of well-
integrated structural knowledge as opposed to the memorization of fragmentary,
unintegrated facts and externalize the conceptual knowledge (both correct and erroneous)
that learners hold in a knowledge domain. Concept maps are a kind of mental models
based on a graph with labeled nodes corresponding to concepts in a problem domain and
with arcs indicating relationships between pairs of concepts. Arcs can be directed or
undirected and with or without linking phrases on them. A linking phrase specifies the
kind of a relationship between concepts. A semantic unit of a concept map is a
proposition. Propositions are concept-link-concept triples which are meaningful
statements about some object or event in the problem domain (Cañas, 2003). Concept
map based tasks can be divided in 1) “fill-in” tasks, where the structure of the concept
map is given to the learner and he/she must fill it using the provided set of concepts
and/or linking phrases, and 2) “construct-a-map” tasks, where the learner must decide on
the structure of the concept map and its content by him/herself.
In the context of the developed system both mentioned types of tasks are provided.
Two kinds of relationships are used: 1) important relationships which show that
relationships between the corresponding concepts are considered as important knowledge
in the learning course, and 2) less important relationships which specify desirable
knowledge. Arcs are directed and linking phrases are provided on them depending on the
degree of task difficulty. Concepts are divided in 1) initial concepts which serve as a
starting point for the learner in the filling or creation of the concept map, and 2) concepts,
which the learner must insert or relate by him/herself.
The system is used in the following way. The teacher defines stages of knowledge
assessment and creates concept maps for all of them. The process of the creation of a
concept map consists from the specification of relevant concepts and relationships among
them. Moreover, the concept map for each stage is nothing else then an extension of the
previous one because new concepts and relationships are added at each stage. Thus, the
concept map of the last stage includes all concepts and relationships among them.
Teacher's created concept maps serve as a standard against which the learners’ concept
maps are compared. During knowledge assessment the learner solves a concept-map
based task corresponding to the assessment stage. After the learner has submitted his/her
solution, the system compares the concept maps of the learner and the teacher, calculates
the score of the learner’s result, gathers statistical information and generates feedback
which is delivered back to the learner.
The system offers five concept-map based tasks, which are ranged from the easiest
to the most difficult (Table 1) taking into account information given to the learner and
workload needed to complete the task (Anohina et al., 2007). Eight transitions between
The 3
rd
International Conference on Virtual Learning, ICVL 2008

83
tasks are implemented allowing the learner to find a task which is the most suitable for
his/her knowledge level. Four transitions increase the degree of task difficulty. They are
carried out after the analysis of the learner’s solution, taking into account whether the
learner has reached the teacher's specified number of points in the current assessment
stage without reducing the degree of difficulty of the original task. So, this is a system’s
adaptive reaction to the learner’s behavior. Other four transitions reduce the degree of
task difficulty and they are carried out by the voluntary request from the learner during
the solving of the task.

Table 1
Tasks Offered in the Concept Map Based Knowledge Assessment System

The type
of the task
The degree of task difficulty
The structure of
a concept map
Linking phrases Concepts
1 Is given
Inserted into the
structure
Must be inserted by the
learner
2 Is given Not used
Must be inserted by the
learner
Fill-in
3 Is given
Must be inserted
by the learner
Must be inserted by the
learner
4 Not given Not used
Must be related by the
learner Construct-
a- map
5
The easiest







The most
difficult
Not given
Must be inserted
by the learner
Must be related by the
learner

An algorithm has been developed for the comparison of learner’s and teacher’s
concept maps (Anohina et al., 2007). It is not based only on the isomorphism of both
graphs, but is sensitive to the arrangement and coherence of concepts taking into account
such aspects as existence of a relationship, locations of both concepts, type and direction
of a relationship, correctness of a linking phrase, etc.
Thus, the system supports knowledge self-assessment as it makes an analysis and
evaluation of learners' concept maps, as well as provides feedback about the learner's
errors. It promotes systematic knowledge assessment because it allows the extension of
the initially created concept map for other assessment stages. Moreover, statistical
information about differences between learners' concept maps and teacher's concept map
is collected providing opportunities for the teacher to improve the learning course.


3. Architecture of the System

The system is implemented as a Web-based application which has three-tier client-
server architecture (Lukashenko et al., 2008). It has the following architectural layers
(Figure 1): 1) a data storage layer, which is represented by Data Base Management
System (DBMS); 2) an application logics layer, which is composed of two parts: the
application server and the server side code running on it; a special persistence and query
framework is used to communicate with the DBMS; and 3) the representation layer or
graphical user interface (GUI).
University of Bucharest and Ovidius University of Constanta


84
Logical principles of the system are based on the Model-View-Controller (MVC)
pattern (eNode, 2002). The representation layer of the system is responsible not only for
the displaying of data, which is the task of the View part, but also it acts as Model. View
part consists of various GUI components (buttons, text input fields, combo boxes lists,
etc.). View handles any events generated by the user, for instance, clicking on a button,
and redirects it to Model, which is located in a separate logical piece. Model, in its turn,
collects data from GUI controls and interacts with Controller (server) by sending and
receiving data and remotely invoking server’s services-methods, which signature is
available on the client side. An action of Model depends on an event, which come from
View. The representation layer is build using java open source graphical user interface
library called swing. JGraph library is used for creation of concept maps. JGoodies library
is included for building more complex GUI layouts.



Figure 1. The Three-Tier Architecture of the System

The application logics layer is implemented as a controller for the entire
application. Apache Tomcat is chosen as an application server. It is a container of
servlets. A servlet is a Java interface, which could be launched by Web server. Servlets
receive clients’ requests and respond to them, usually across HyperText Transfer
Protocol. There is a basic implementation of this interface (for example HttpServlet), but
it can be extended by creating user’s defined event handlers and data transformation for
concrete business logic.
The application server receives remote calls from the client and redirects them to
the appropriate servlet. The information about a servlet is included in the remote call. The
servlet handles a call and launches the appropriate method needed for the communication
with the database or for the execution of business logic. The application server does not
use SQL queries to perform data manipulations. Instead of that, Java object oriented
framework, namely Hibernate, is used.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

85
Hibernate is a high performance object/relational persistence and query framework.
It allows programmers to develop persistent classes following object-oriented paradigm,
including associations, inheritance, polymorphism, composition and collections.
Hibernate provides opportunities to create queries by using SQL extension HQL, native
SQL, or object-oriented criteria (Red Hat, 2006).
To start working with Hibernate, it is necessary to define two major things: an
entity and its xml mapping or annotation. The file “hibernate.properties” with the
extension .xml describes a basic configuration, that is, how and with which database
Hibernate will work: URL of a database, a user name, a password, presented entities, and
so on. An entity represents a real world object with the set of simple attributes. Xml
mapping represents this entity as a relational table in the database, and describes metadata
of entity’s attributes and relations with other entities. Entities can reference each other,
can have child collections of other entities, and so on. The structure and relations of an
entity might be as complex and sophisticated, as it is needed for the modeling of real
world objects.
Hibernate provides handy and flexible API for any manipulations with persistent
entities. So a programmer works with DBMS via Hibernate as with Java classes (tables)
and objects (rows). The one of the most powerful feature of Hibernate is lazy loading.
Lazy loading is a mechanism for comfortable work with large amounts of data even
without loading them into computer memory, excepting cases, when it is necessary to
perform some actions with a definite piece of data (Red Hat, 2006).
There is one more thing to add about communication between Hibernate and
DBMS. The framework performs any loading/saving/updating operations with data using
its own generated SQL, because DBMS “understands” only this language.
For the implementation of the data storage layer the Data Base Management
System Postgresql is chosen. This software is open source and supports PL/SQL.
As it is shown in Figure 1 the concept map based knowledge assessment system
can be divided into three logical domains: administrator, teacher and student. Each
domain has its own goal, but they are strictly linked together. Functionality of each
domain can be used by one of three user roles which names correspond to the names of
the domains. An administrator is responsible for the administration and maintenance of
the whole system using such functions as input, editing and deleting of data about users
(teachers and students), courses and student groups. Teacher domain provides all
necessary functions for the creation of concept maps for any course and defining of their
attributes, as well as for the viewing of learners’ results. Functionality of the student
domain includes all things related to the completion of the concept map based tasks by
learners and providing of feedback after the completion of the task.


4. Example of the System’s Operation

First of all the teacher creates concept maps for chosen stages of knowledge
assessment by defining relevant concepts and relationships among them. Figure 2
displays the partly created teacher’s concept map for the first assessment stage and a
University of Bucharest and Ovidius University of Constanta


86
dialogue window where data of a new concept must be provided. Only one concept, that
is, New Year, is defined as an initial concept and its color differs from the color of other
concepts. Three relationships (marked by thick line) are important relationships, and one
marked by thin line is a less important relationship.
After the creation of the concept map the teacher must define the publication data
of the concept map, the initial degree of task difficulty, the relative number of points
needed to move the learner to the higher degree of task difficulty and time for the
completion of the task (if necessary). Let’s consider that the initial degree of task
difficulty is the fourth degree, the relative number of points is 75%, and time for the task
completion is not provided.
The concept map presented to learners at the first assessment stage is displayed in
Figure 3. During the solving of the task of the fourth degree of difficulty learners must
create their own concept maps using the offered set of concepts. The technique of drag-
and-drop must be used to move concepts from the concept palette to the working space.
In order to relate concepts two buttons are provided at the top of the window.
Assume that one of learners has related all concepts and submitted his/her solution
without reducing the degree of task difficulty. Figure 4 shows the learner’s created
concept map and an example of feedback provided for each relationship. Different colors
are used to display different degrees of correctness of relationships.


Figure 2. The Teacher’s Created Concept Map for the First Assessment Stage
The 3
rd
International Conference on Virtual Learning, ICVL 2008

87



Figure 3. The Task of the Fourth Degree of Difficulty

Now, assume that another learner had partly created his/her concept map at the
fourth degree of task difficulty and after that asked the system to reduce the degree of
task difficulty. The learner receives the task of the third degree of task difficultly where
the structure of the concept map is given and it is necessary to insert given concepts and
to provide linking phrases (Figure 5). However, all propositions created by the learner at
the fourth degree of task difficulty are kept. After the completion of the task this learner
will receive the same form of feedback which is presented in Figure 4, but at the next
assessment stage she/he will start to solve the task on the third degree of task difficulty
because of the reduction of task difficulty of the original task.



Figure 4. Feedback Provided After the Completion of the Task
University of Bucharest and Ovidius University of Constanta


88



Figure 5. The Task of the Third Degree of Difficulty


5. Learner’s support

The knowledge assessment system has been experimentally evaluated in 7 learning
courses with participation of 149 students. The results show that students positively
evaluate the functionality and the user interface of the system. However, a part of
students (about 60%) experience difficulties during the solving of the concept map based
tasks. One of the reasons mentioned by students is the lack of learner’s support. Thus, the
current directions in the development of the system are related to the implementation of
different kinds of learner’s support especially help and feedback. The purpose of help is
to balance the degree of task difficulty and learner’s knowledge level in order to help the
learner to complete the task. In turn, feedback is aimed to give the learner information
about the correctness of his/her actions and progress towards the goal, that is, towards the
successful completion of the task.
Three kinds of learner’s support (Table 2) are chosen for further implementation. In
addition, there plans to use a student model for the provision of such type of explanations
which are preferred by the learner or which the system recognizes as the most suitable for
the learner taking into account learner’s characteristics. This will make support more
adaptive increasing its overall usefulness.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

89
Table 2
Kinds of Learner’s Support

Description of a support form
Kind of
support
Helping
nature
Tutoring
nature
The system inserts the learner’s selected concept into
the right node within the structure of the concept map,
thus, decreasing the number of concepts which the
learner must insert by him/herself
Help + –
The system gives an explanation (definition of the
concept, short description or example) of the learner’s
selected concept helping the learner to understand the
concept and its relations with other concepts
Help + +
The system checks the correctness of the learner’s
created proposition and in case of its incorrectness
provides appropriate explanations of both concepts
involved in the proposition
Feedback + +


6. Conclusions

The paper presents the architecture and working principles of the concept map
based knowledge assessment system developed at Riga Technical University. The main
components and technologies used for the implementation of the system are described
and an example of the system operation is demonstrated. Despite of fact that the system
has already reached the certain level of maturity and has been used successfully in
practice authors continue to improve its functionality. One of the significant development
directions is the implementation of different kinds of learner’s support inter alia adaptive
mechanisms of help and feedback on the basis of a student model.



REFERENCES


ANOHINA, A., POZDNAKOVS, D. and GRUNDSPENKIS, J. (2007), Changing the Degree of
Task Difficulty in Concept Map Based Assessment System. In Proceedings of the IADIS
International Conference “e-Learning 2007”, Lisbon, Portugal, 443-450.
CAÑAS, A. J. (2003), A Summary of Literature Pertaining to the Use of Concept Mapping
Techniques and Technologies for Education and Performance Support. Technical report:
Pensacola, FL.
eNode, Inc. (2002), Model-View-Controller Pattern, http://www.enode.com/x/markup/
tutorial/mvc.html
University of Bucharest and Ovidius University of Constanta


90
LUKASHENKO, R., VILKELIS, M. and ANOHINA, A. (2008), Deciding on the Architecture of
the Concept Map Based Knowledge Assessment System. In Proceedings of the International
Conference on Computer Systems and Technologies, Gabrovo, Bulgaria (in print).
Red Hat, Inc (2006), HIBERNATE – Relational Persistence for Idiomatic Java,
http://www.hibernate.org/hib_docs/reference/en/html/tutorial.html



The 3
rd
International Conference on Virtual Learning, ICVL 2008

91
Measurement and Control of Statistics Learning Processes based
on Constructivist Feedback and Reproducible Computing

Patrick Wessa

K.U.Leuven Association, Lessius Dept. of Business Studies, Belgium
E-mail: patrick@wessa.net


Abstract
This article introduces a new approach to statistics education that allows us to
accurately measure and control key aspects of the computations and communication
processes that are involved in non-rote learning within the pedagogical paradigm of
Constructivism. The solution that is presented relies on a newly developed
technology (hosted at http://www.freestatistics.org) and computing framework
(hosted at http://www.wessa.net) that supports reproducibility and reusability of
statistical research results that are presented in a so-called Compendium.
Reproducible computing leads to responsible learning behaviour, and a stream of
high-quality communications that emerges when students are engaged in peer
review activities. More importantly, the proposed solution provides a series of
objective measurements of actual learning processes that are otherwise
unobservable. A comparison between actual and reported data, demonstrates that
reported learning process measurements are highly misleading in unexpected ways.
However, reproducible computing and objective measurements of actual learning
behaviour, reveal important guidelines that allow us to improve the effectiveness of
learning and the e-learning system.

Keywords: Reproducible Computing, Virtual Learning Environment, Communication,
Statistics Education, Learning Processes.


1. Introduction and Literature

In education-related research it is common practice to investigate learning
processes through measurements that are based on questionnaires. Reported measures
often reveal interesting information about a wide variety of aspects of computing-assisted
learning such as: computer attitudes (Meelissen and Drent, 2008); computer emotions and
knowledge (Kay 2008); learner experiences and satisfaction (Sun et al. 2008); etc. The
importance of such measurements has been highlighted by many authors from various
perspectives (Chen, 2008; Hilton et al., 2004; Galotti et al., 1999) – especially from the
perspective of the constructivist pedagogical paradigm (Von Glasersfeld, 1987; Smith,
1999; Eggen et al., 2001; Mvududu, 2003).
These reported measures, while intrinsically interesting, may not always provide us
with the information we need to assess and improve systems that support e-learning.
Moreover, the implementation of new learning technologies and data analysis tools opens
up a wide array of measurement opportunities which leads to new areas of research. An
University of Bucharest and Ovidius University of Constanta


92
excellent example is the use of data mining tools in the open source e-learning
environment called Moodle (Romero et al., 2008).
Even though it seems to be very difficult to measure and empirically prove
(O’Dwyer et al., 2008), there is no doubt in my mind that the introduction of computers in
homes and classrooms has led to an improvement in overall learning productivity,
educational communication mechanisms, social constructivism, and collaboration.
However, the use of computers and software in statistics education may – unwillingly –
result in several types of adverse effects because the complex processes that are required
to learn and (truly) understand statistical concepts are often mystified by technicalities
and a variety of practical problems that have nothing to do with mathematics or statistics.
It is within this context that I argue that a system for Quality Control should be embedded
into the e-learning environment which is not limited to the Virtual Learning Environment
but extends to the statistical software, databases, and learning repositories.
There is an important, additional benefit for implementing such a monitoring and
control system – it is directly related to the problem of irreproducible research which has
received a great deal of attention within the statistical computing community (de Leeuw,
2001; Peng et al., 2006; Schwab et al., 2000; Green, 2003; Gentleman, 2005; Koenker
and Zeileis, 2007; Donoho and Huo, 2004). The most prominent citation about the
problem of irreproducible research is called Claerbout's principle: An article about
computational science in a scientific publication is not the scholarship itself, it is merely
advertising of the scholarship. The actual scholarship is the complete software
development environment and that complete set of instructions that generated the figures.
(de Leeuw, 2001).
Several solutions have been proposed (Buckheit and Donoho, 1995; Donoho and
Huo, 2004; Leisch, 2003) but have not been adopted in statistics education because they
require students to understand the technicalities of scientific word processing (LaTex) and
statistical programming (R code). Based on a newly developed e-learning environment I
propose a solution that is feasible for educational purposes and allows us to monitor,
research, and control the learning processes based on the dynamics of between-student
communication and collaboration.


2. Reproducible Computing
2.1. R Framework and Compendium Platform

The R Framework allows educators and scientists to develop new, tailor-made
statistical software (based on the R language) within the context of an open-access
business model that allows us to create, disseminate, and maintain software modules
efficiently and with a very low cost in terms of computing resources and maintenance
efforts (Wessa, 2008a). The so-called R modules empower students to perform statistical
analysis through a web-based interface that does not require them to download or install
anything on the client machine. This permits students to focus primarily on the
interpretation of the analysis – however, the R Framework also allows advanced students
and scientists to inspect and change the R code that was coded by the original author.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

93
This results in the creation of so-called “derived” R modules that may be better suited for
particular purposes.
If a derived R module contains generic improvements or if a computation needs to
be communicated to other students/scientists then it is necessary to have a simple,
transparent mechanism that allows one to permanently store the computation in a
repository of computational objects that can be easily retrieved, recomputed, and reused.
Such a repository was recently created within the OOF 2007/13 project of the
K.U.Leuven Association and is called the Compendium Platform. The main reason for
creating the R Framework and the Compendium Platform, is that it allows anyone to
create and use Compendia of reproducible research. A Compendium is defined as (Wessa
2008b): a research document where each computation is referenced by a unique URL
that points to an object that contains all the information that is necessary to recompute it.
Such documents can be easily created (even by students) and permit any reader to
(exactly) recompute the statistical results that are presented therein. A few simple clicks
are sufficient to have the R Framework reproduce the results and to reuse them in derived
work (Wessa 2008b). The practical implications of this technology will become obvious
in section 3 because the three figures that are presented can be recomputed and reused
through the Compendium Platform.

2.2. Communication, Feedback, and Learning

The concept of Reproducible Computing was implemented in several undergraduate
statistics courses in order to test the new system and to measure key aspects of the
educational activities and experiences. Two different student populations were investigated
in detail: a group of (academic) bachelor students, and a group of so-called “switching”
students. The second population is of particular interest because it consists of students
who obtained a (professional) bachelor degree and decided to make the “switch” to an
academic master which requires them to complete a preparatory year.
On the one hand, switching students are highly motivated and more mature than the
bachelor students. A priori, one would expect them to prefer practical activities (such as
communication and computing) above theory and critical reflection. On the other hand,
one might expect the bachelor students to have a more critical (scientific) attitude and
better mathematical background than the switching students.
Students from both populations took a similar statistics course which covered
topics from introductory statistics, regression analysis, and introductory time series
analysis. The main learning activities in both statistics courses were based on a weekly
series of workshops where each student was required to investigate practical, empirical
problems. At the end of each week, students submitted their papers electronically. During
the lectures I proposed a series of solutions and illustrated commonly made mistakes.
After the lectures, students had to work on the next assignment and complete a series of
peer reviews (assessments) about the work that was submitted the week before. The
assessment grades did not count towards the final score – however, each submitted peer
review was accompanied by verbal feedback messages. I graded a random sample of
these messages in order to provide students with an incentive to take the review process
University of Bucharest and Ovidius University of Constanta


94
seriously. There is strong empirical evidence that this approach had beneficial effects on
non-rote learning of statistical concepts (Wessa 2008c).


3. Measurement and Control

In Wessa (2008b) it is illustrated how the Compendium Platform's repository
supports quality control of the statistical software and accompanying documentation for
students. On the one hand, reproducible computing allows students to accurately
communicate computational problems and questions without the need to understand the
underlying technicalities. On the other hand, it allows the educator (and creator of the
computational software) to analyse the reported problem (based on the detailed, raw
output of the R engine that executed the request) and to transparently communicate the
solutions to the students. In addition, the measurement of learning activities and
experiences is a conditio sine qua non for controlling the overall quality of learning
systems. This will be illustrated based on the data that was collected for both student
groups. At the same time, the importance of objective (as opposed to reported)
measurements is illustrated based on a simple, comparative diagnostic tool.
The reported measurements were obtained through questionnaires on a 5-point
Likert scale and should consequently be treated as ordinal data. The questions were based
on well-known psychological surveys (Galotti et al. 1999; COLLES 2004) and an
extended version of the IBM computer system usability survey (Lewis 1993). Useful data
was obtained from a total of 111 bachelor students and 129 switching students – the
response ratio was very high (between 82.9% and 92% depending on the questionnaire).
All observations of actual learning activities were measured on a ratio scale (the number
of archived computations and the number of submitted feedback messages). A total
number of 34438 meaningful, verbal feedback communications and 6587 archived
computations were registered. In order to compare the actual and reported data, all
measurements were converted to ordinal rank orders. In addition, the Pearson's rho
correlations and Kendall's tau rank correlations (Arndt et al. 1999; Arnd, Magnotta 2001)
that represent the degree of linear association between the properties under investigation
were computed (these can be consulted in the archived computations about the Figures).
In electronic versions of this paper, one can simply (ctrl-)click the pictures to view the
archived computation in the repository. Readers of the printed version of this document,
are referred to the bibliography where three references can be found (including the URLs)
about the statistical computations that have been stored at www.freestatistics.org.
Figure 1 displays the bivariate kernel density (Lucy et al. 2002) between the rank
order of the number of feedback messages that have been submitted in peer reviews (x-
axis) and the rank order of the number of (reproducible) computations that have been
archived in the repository (y-axis). The rank orders have been computed within the
Bachelor population for the top panels, and within the Switching population for the
bottom panels. This implies that the ranks that are attributed to female and male students
are expressed on the same axes and can be compared. Figure 1 clearly demonstrates that
female bachelor students are much more involved in feedback and computing than their
The 3
rd
International Conference on Virtual Learning, ICVL 2008

95
male colleagues. At the same time, female switching students are more computing-
oriented whereas the male switching students seem to have a slight preference for
feedback communication. This information has important repercussions for controlling
the quality of the learning environment and it provides clear guidelines towards actions
that should be taken (by me) to improve participatory incentives towards male bachelor
students in future courses. Would I have been able to gain this insight based on reported
measurements alone? The answer is clearly negative (as is illustrated in Figures 2 and 3).



Figure 1. Submitted Feedback versus Reproducible Computations

It is quite obvious that male bachelor students highly over-estimate their
performance in terms of feedback submissions (see Figure 2) because the rank orders of
reported measures (x-axis) are higher than the ranks of actual feedback submissions (y-
axis). Female bachelor students however, underestimate their involvement (relative to
their male colleagues) because they are concentrated above the diagonal line. In the male
switching student population several clusters of high density can be detected which leads
us to conclude that we cannot treat them as one homogeneous group.
In Figure 3 the comparison between reported computing measures (x-axis) and
actual computing (y-axis) leads to similar conclusions. Male bachelor students highly
exaggerate their efforts, whereas female bachelor and switching students underestimate
themselves. The group of male switching students is heterogeneous.



University of Bucharest and Ovidius University of Constanta


96


Figure 2. Reported versus Actual Submitted Feedback



Figure 3. Reported versus Actual Reproducible Computing
The 3
rd
International Conference on Virtual Learning, ICVL 2008

97
Overall, the testimony of students is extremely misleading and poorly correlated with
actual observations. If we would have recomputed Figure 1 with reported measures then the
conclusions would have been the opposite of what is true. The reader can try out this experiment
by simply using the online R module at http://www.wessa.net/rwasp_icvl2008.wasp
(specify a picture width and height of 800 for best results).
The good news is that we now have a tool available to assess actual and reported
learning activities for any student population that makes use of the new compendium
technology. Ultimately, this allows us to take control and improve the e-learning
environment, the statistical software, the course materials, and overall learning
experiences of all students.



REFERENCES


ARNDT, S., TURVEY, C., ANDREASEN, N. (1999), Correlating and predicting psychiatric
symptom ratings: Spearman’s r versus Kendall’s tau correlation, Journal of Psychiatric
Research, 33, 97-104.
ARNDT, S., MAGNOTTA, V. (2001), Generating random series with known values of Kendall’s
tau, Computer Methods and Programs in Biomedicine, 65, 17-23.
Attitudes to Thinking and Learning Survey, (n.d.), Retrieved December 22, 2004, from
http://www.moodle.org/
BENSON, J. (1989), Structural components of statistical test anxiety in adults: An exploratory
study, Journal of Experimental Education, 57, 247-261.
CHAMBERS, J. M., CLEVELAND, W. S., KLEINER, B. and TUKEY, P. A. (1983), Graphical
Methods for Data Analysis., Wadsworth & Brooks/Cole.
CHEN, Z. (2008), Learning about Learners: System Learning in Virtual Learning Environment, International
Journal of Computers, Communications & Control, Vol. III (2008), No. 1, pp. 33-40.
COLLES (2004), Constructivist On-Line Learning Environment Survey, Retrieved December 22,
2004, from http://www.moodle.org/
EGGEN, P., and KAUCHAK, D. (2001), Educational Psychology: Windows on Classrooms (5th
ed.), Upper Saddle River, NJ: Prentice Hall.
GALOTTI, K. M., CLINCHY, B. M., AINSWORTH, K., LAVIN, B. & MANSFIELD, A. F. (1999), A new
way of assessing ways of knowing: the attitudes towards thinking and learning survey
(ATTLS), Sex roles 40(9/10) p745-766.
HILTON, S., SCHAU, C., OLSEN, J. (2004), Survey of Attitudes Toward Statistics: Factor Structure
Invariance by Gender and by Administration Time, Structural Equation Modeling, Volume 11,
Number 1.
KAY, R. H. (2008), Exploring the relationship between emotions and the acquisition of computer
knowledge, Computers & Education, 50, 1269-1283.
LEWIS, J. R. (1993), IBM Computer Usability Satisfaction Questionnaires: Psychometric
Evaluation and Instructions for Use, IBM Corporation, Technical Report 54.786.
LUCY, D., AYKROYD, R. G. & POLLARD, A. M. (2002), Non-parametric calibration for age
estimation. Applied Statistics 51(2): 183-196.
MEELISSEN M. R. M, DRENT M. (2008), Gender diferences in computer attitudes: Does the
school matter?, Computers in Human Behavior, 24, 969-985.
University of Bucharest and Ovidius University of Constanta


98
MILLER, JACQUELINE B., Examining the interplay between constructivism and different learning
styles, Retrieved October 20, 2005 from www.stat.auckland.ac.nz/~iase/publications/
1/8a4_mill.pdf
MVUDUDU, NYARADZO (2003), A Cross-Cultural Study of the Connection Between Students'
Attitudes Toward Statistics and the Use of Constructivist Strategies in the Course, Journal
of Statistics Education, Volume 11, Number 3.
O’DWYER, L. M., RUSSELL, M., BEBELL, D., SEELEY, K. (2008), Examining the Relationship
between Students’ Mathematics Test Scores and Computer Use at Home and at School,
Journal of Technology, Learning, and Assessment, 6 (5).
ROMERO, C., VENTURA, S., GARCÍA, E. (2008), Data mining in course management systems:
Moodle case study and tutorial, Computers & Education, 51, 368-384.
SMITH, ERICK (1999), Social Constructivism, Individual Constructivism and the Role of Computers
in Mathematics Education, Journal of mathematical behavior, Volume 17, Number 4.
Statistical Computations at FreeStatistics.org (2008a), Office for Research Development and Education,
URL http://www.freestatistics.org/blog/date/2008/Jun/30/t1214840420q0fyankop4x9ebf.htm
Statistical Computations at FreeStatistics.org (2008b), Office for Research Development and Education,
URL http://www.freestatistics.org/blog/date/2008/Jun/30/t12148409608o0dnj2k4s04jil.htm
Statistical Computations at FreeStatistics.org (2008c), Office for Research Development and Education,
URL http://www.freestatistics.org/blog/date/2008/Jun/30/t1214841152sn6jlyhgseclgqm.htm
SUN, P., TSAI, R. J., FINGER, G., CHEN, Y.,YEH, D. (2008), What drives a successful e-Learning?
An empirical investigation of the critical factors influencing learner satisfaction, Computers
& Education, 50, 1183-1202.
VON GLASERSFELD, E. (1987), “Learning as a Constructive Activity”, in Problems of Representation,
in the Teaching and Learning of Mathematics, Hillsdale, NJ: Lawrence Erlbaum Associates, 3-17.
WESSA, P. (2008a), A framework for statistical software development, maintenance, and publishing
within an open-access business model, Computational Statistics, www.springerlink.com
(DOI 10.1007/s00180-008-0107-y)
WESSA, P. (2008b), Learning Statistics based on the Compendium and Reproducible Computing,
submitted to be presented and published in Proceedings of the International Conference on
Education and Information Technology (ICEIT'08), San Francisco, USA.
WESSA, P. (2008c), How Reproducible Research Leads to Non-Rote Learning Within a Socially
Constructivist E-Learning Environment, submitted to be presented and published in
Proceedings of the 7th European Conference on e-Learning (ECEL'08), Cyprus.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

99









Section

MODELS & METHODOLOGIES




• Innovative Teaching and Learning Technologies
• Web-based Methods and Tools in Traditional, Online
Education and Training
• Collaborative E-Learning, E-Pedagogy,
• Design and Development of Online Courseware
• Information and Knowledge Processing
• Knowledge Representation and Ontologism
• Cognitive Modelling and Intelligent systems
• Algorithms and Programming for Modelling




The 3
rd
International Conference on Virtual Learning, ICVL 2008

101
Development of Group Division Algorithm And Discussion
Support System for Intra-class Discussions

Ikuo Kitagaki

Research Institute for Higher Education, Hiroshima University
2-12-1 Kagamiyama, Higashi-hiroshima, 739-8512, Japan
E-mail: kitagaki@hiroshima-u.ac.jp


Abstract
I studied the computer system that was characterized by the algorithm for dividing one
class of students into several groups for group discussion. This paper describes the
configuration of the proposed computer system, the algorithm of the group division, and
the execution process of actual group discussions, assisted by this system, about specific
topics.

Keywords: Group division, Discussion support system, Algorithm


1. Introduction

With the development of e-learning systems, computer-assisted collaborative
learning and group learning have become more popular. I studied the computer system
that was characterized by the algorithm for dividing one class of students into several
groups for group discussion(Akahori, 1997). This computer system is used for the group
division, however, it is not used in the later group discussions because the actual
discussions are made as traditional face-to-face communication activities. This computer
system can be also called the computer-assisted, group discussion support system.
Concretely speaking, this system divides one class into multiple groups according to
students’ answers of a discussion topic by using a specific algorithm, and each student is
notified of the names of the members who belong to the same group. The students then
form groups according to the notified information, exchange opinions, and discuss the
topic to increase understanding. As almost all students have their own cell-phones, the
computer server collects information necessary from students and distributes information
to students via cell-phones.
This paper describes the configuration of the above computer system, and the
algorithm of the group division. This paper also describes the execution process of actual
group discussions, assisted by this system, about specific topics.
There is a difference in the group division algorithm between this paper and the
previous paper(kitagaki et.al., 2007). In the previous paper, we divided a student class
based on the students’ answers to the test. That is, values 1 and 0 were assigned to the
right and wrong answers of the test respectively, and the class was divided into groups
according to these assigned values. On the other hand, in this new paper, we divided a
University of Bucharest and Ovidius University of Constanta


102
student class into groups according to the information on debater’s choices about the
discussion topic instead of the above test answers. In the group dividing process, as the
similarity of contents among these choices should be considered, we assigned values 0
and 1 to the choices, respectively.


2. Discussion Support System

Discussion support system is implementing the system flow shown in Figure 1.
(1)Registration of student’s attributes: The mail address of students, name and other
information such as sex, generation can be inputted to the computer… Among these, mail
address is necessary and their names are used for students to know all the group member.
These registrations are done on a Web page. The URL of the page is informed to all
students in advance.



(2)Sending URL of a topic and its choices: The teacher selects a subject among
prepared topics. Then the computer sends the URL for browsing the topic to all discussants.
server
(incl. local PC)

Student attribute
list





D
i
s
c
u
s
s
a
n
t

s

c
e
l
l

p
h
o
n
e


Topic of discuss.

Assembl.
Ans.data

Group division


Inform. of
group member
et.al.

Regist. thru.Web
Transmit. URL

Brows. Web,
transmit. answer
p
a
r
a
m
e
t
e
r

(4)

Member, ans.data
(1)
(2)
(3)
(5)
Figure 1. System configuration
The 3
rd
International Conference on Virtual Learning, ICVL 2008

103
(3)Browzing the Web of a subject and sending the student’s answer to the server:
All students make an access to the URL mentioned above, then read the topic and the
choices. They select a choice among the choices for the given topic then send it to the
server. All the answer are gathered and stored in the server.
(4)group division: The computer server makes the group division according to the
student’s answer as transaction data. The basic idea of the group division is explained in
the next section. In the actual administration, the following parameters and the necessary
information ought to be inputted prior to the group division.
1. the value with which a topic is discussed.
2. the difference of choices in their contents.
3. the number of a group constituents
When the information of group division is obtained as a processed result, it is
possible for a teacher to add, to it, remarks to each group, remarks to each individual and
remarks to the all.
(5)sending the group division information to the students: The computer server sends
to each student’s mail address the group members, each answer and the remarks above if
any.
Through the process above, each student is informed the name of all member which
belong to the same group. Then, each group gathers somewhere inside or outside of the
classroom and starts to deeply consider the topic by discussion.


3. Method of group division

Group division can be made by two kinds of criterion as the followings.
(a) difference: Groups are made so that choices of each member may be different
from those of others as much as possible.
(b) similarity: Group are made so that choices of each member are similar with
those of others as much as possible.
Two criteria are reverse in their evaluation of ‘goodness’. Thus it is enough only to
explain criterion (a). As for the criterion (a), two methods have been proposed (Kitagaki,
1996; Kitagaki et. al., 1981). The proposed system in this material adopts the simpler
method (Kitagaki et. al., 1981). The argorithm is outlined below.

topic sets: M
topic:
) ( M mi ∈

value of the topic: ) ( i m v
group sets: G
relevant group: ) ( G g ∈
bigness of group ‘g’: |g|
discussant(student): ) ( g xj ∈
selected choice: a(m
i
,x
j
)
difference of choices selected by discussant x
j
and discussant x
k
: d{a(m
i
,x
j
),a(m
i
,x
k
)}
University of Bucharest and Ovidius University of Constanta


104
goodness of group division using criterion (a): α
g

goodness of group division using criterion (b): α’
g

goodness per capita of group division: β

In the definition above, both of ‘value of the topic’ and ’’difference of choices
selected by discussant x
j
and discussant x
k
’ are the value between 0 and 1. The ‘bigness
of group ‘g’’ is the number of a group. The number is not always the same for all group.
But its algorithm is abbreviated here. All the said variables have to be determined in
advance. The ‘goodness per capita of group division β’ is determined as the following.

[1]
U
k M m g x
k i j i
i j
i g x m a x m a d m v )} , ( ), , ( { ) (
∑ ∑
∈ ∈
= α
[2]
∑ ∑
∈ ∈
=
G g G g
g g | | / α β

In the equation [2], group division which makes the value maximum is the optimal
solution. In order to get the optimum, however, it is necessary to administrate the calculation for
all the combination of groups. It is actually difficult to get it because of time for its calculation.
Thus a simple method is implemented (Kitagaki et. al., 1980) as the following. Its example
deals with the case that thirty discussants are divided into ten groups with three discussants each.
As the initial status, I suppose that the computer fix the discussants x
1
, …, x
30
as
shown in ‘n = 1’ Figure 2, and define the value of β in eq.[2] as β
1
. Then I let it compare
cell1 with each cell thereafter one by one. First, let it exchange cell1 x
1
for cell2 x
2
to
obtain the pattern as shown in ‘n = 2’ then get the value β as β
1,2
. It is clear that β
1,2
is
same as β
1
in their value. Thus there is no reason to exchange thus it ought to be
withdrawn. Second, it is obvious that the exchange of x
1
and x
3
leads to the same result
as above. It is cell1 and cell4 that has actual meaning of exchange because they belong to
different groups in the initial pattern. If β
1
is bigger than(or equal to) β
1,4
, the computer
regards the pattern of ‘n = 4’ as not better pattern than the one of ‘n = 1’ then the
exchange ought to be withdrawn. On the other hand, if β
1
is smaller than β
1,4
, it regards
the pattern of ‘n = 4’ better than the one of ‘n = 1’ then the exchange ought to be done to
get the new pattern. Based upon the new pattern, it searches for a better pattern. The
search for the better pattern is succeeded in the same way.
Consequently, the exchange of two cells are done in the following order, and as a result,
the number of exchange becomes 870 ( = 29*30) in all. (Actually the exchange of two
cells in a group ought to be omitted.)

cell1 and cell2, cell1 and cell3, cell1 and cell4,cell1 and cell5, …,cell1 and
cell29, cell1 and cell30
cell2 and cell3, cell2 and cell4, cell2and cell5, …,cell2 and
cell29, cell2 and cell30
cell3 and cell4, cell3 and cell5, …,cell3 and
cell29, cell3 and cell30
………………………………………………
cell29 and cell30
The 3
rd
International Conference on Virtual Learning, ICVL 2008

105

Supposing the number of discussant to be ‘n’, the number of the said exchange
becomes ‘n (n – 1)’. For each exchange, the computer gets the value of β, then the
optimal group division is obtained.

n=1
cell1 cell2 cell3 cell4 cell5 cell6 … cell28 cell29 cell30

x
1
x
2
x
3
x
4
x
5
x
6
… x
28
x
29
x
30
n=2
cell1 cell2 cell3 cell4 cell5 cell6 … cell28 cell29 cell30

x
2
x
1
x
3
x
4
x
5
x
6
… x
28
x
29
x
30
n=4
cell1 cell2 cell3 cell4 cell5 cell6 … cell28 cell29 cell30

x
4
x
2
x
3
x
1
x
5
x
6
… x
28
x
29
x
30
g
1
g
2
g
10
g
1
g
2
g
10
g
1
g
2
g
10


Figure 2. The exchange of two cells(in the case that a classroom consists of thirty students)

If method (b) shown in the beginning of this section is implemented for group
division criterion, we have only to use eq.[3] instead of eq.[1].

[3]
U
k M m g x
k i j i
i j
i g x m a x m a d m v )}} , ( ), , ( { 1 { ) ( ' − =
∑ ∑
∈ ∈
α


4. The administration of discussion classroom

As a topic for the proposed discussion, I raised a topic of career development to
which most students might be relevant. The topic relates the consideration on the answer
in an interview in job hunting. I administrated the classroom discussion four times. In every
administration, the same topic was used. Two experimental administrations are discussed below.
[experimental 1(E1)] Fifteen Hiroshima University students served as subject
(twelve undergraduate students and three graduate students). Among them, nine students
were science in major., Jul. 26, 2007.
University of Bucharest and Ovidius University of Constanta


106
[experimental 2(E2)] Fourteen Hiroshima University students served as subject
(seven undergraduate students and seven graduate students). Among them, nine students
were science in major., Nov. 11, 2007.
In both of E1 and E2, I used the same room and implemented the same topic which
consists of two rounds, R1 and R2, that is, two ‘Question-Answer’ set. But the students
were different in those two experiments. In either round, the question exemplified in
Table 1 was used. In its actual administration, I let the students inform others in a group
of the choice which each student has selected then discuss what would be a better answer
for the given answer. Lastly I let them make and write a better answer then offer it as a
report with of their consensus.



As most students in a group were not acquainted with each other, it was assumed
that, when the information of a group member were presented in the step of Figure 1(5), it
gets difficult for them to make a group. Thus ID number proper to each group was
informed all the classroom, leading to easier making groups. Addition to that, as group
leaders is necessary in order to facilitate the discussion well, how to determine a leader in
a group was also informed them. Discussion time was set to nearly fifteen minutes, which
was also informed them as a comment.
After each administration of R1 and R2, the following questionnaires have been
examined to all the students.

(Influence of selecting a choice on the discussion flow)
1. I don’t think that the information of a choice selection had an influence on the
relevant discussion flow. In other words, discussion flow must have been the same as the
one without selecting a choice.
2. I think that the information of a choice selection had an influence on the relevant
discussion flow to some extent.
Q: What worker do y ou want to be?
A: As human is not working alone, I want to be a
worker regarding communication with others in the
actual work where I should say whatever necessary
to say and listen to them whenever to do that .
# As your remark for the underlined parts, select
a choice that is nearest to it.
1. The expression is vague and fuzzy thus have
little impact.
2. No fresh awareness as a new comer. can be felt.
3. Everybody can say that thus little impact are there.
4. It is the president of a company to say that.
Thus, if the matter gets worse, it may give them
a feeling of impoliteness.
Table 1
A questionnaire and the answer
(Nakatani, 1995)


The 3
rd
International Conference on Virtual Learning, ICVL 2008

107
3. I think that the information of a choice selection had an influence on the relevant
discussion flow to great extent.

In R1 and R2, presented topics were the same type thus those answers were added to
have been statistically processed altogether. The result is shown in Table 2(a). There is
also shown the result in the case that E1 and E2 were combined in their data.
From the test result shown in the table, the average choice for the questionnaire is
said to be 2 (or in-between 2 and 3). It means that their average remark is that selecting a
choice has an influence on the discussion flow as their awareness.
Besides the experiment shown in this material, I have done the questionnaire survey
comparing criterion (a) with criterion (b) shown in the 3
rd
section supposing the case that groups
were divided according the criterion (b). As their awareness, it got obvious that they felt more
fruitful in their discussion with variety of choices in a group than that with a similar choices.

Table 2
Answer data processing of ‘ influence of selecting a choice’
(a)basic statistics

number of students
selecting a choice
condition

X σ
E1R1&R2 0.30 0.57 0.13 1.90 0.65
E2R1&R2 0.07 0.46 0.46 2.39 0.61
E1&E2,R1&R2 0.19 0.52 0.29 2.10 0.69

Number of the students is fifteen 15 in E1, and fourteen in E2.
X : average, σ : standard deviation

(b) Z-test(using normalized distribution)

m
condition 1 2 3
(E1, R1&R2) – 7.42
*
0.86 9.13
*

(E2, R1&R2) – 11.9
*
– 3.37
*
5.20
*

(E1&E2, R1&R2) – 12.2
*
– 1.15 9.94
*


* Hypothesis ’X=m’ has been rejected(p<0.01),
σ
N m X
Z
) ( −
=


5. Conclusion

In this research, I have developed an algorithm for group division where discussion
on a subject is done in a group. For that objective, I made a system configuration. As a result
University of Bucharest and Ovidius University of Constanta


108
of its administration, it got clear that the discussants have an awareness that letting them
know their selected choice each other gave an influence on the further discussion flow.



REFERENCES


AKAHORI, K. et. al. (1997, in Japanese), The skill of university classroom teaching, Daiichi-
houki, Tokyo, 142-145.
KITAGAKI, I., HIKITA, A.,TAKEYA, M., FUJIHARA, Y. (2007), Development of an algorithm
for groupware modeling for a collaborative learning, Int’l Journal of Computers,
Communication & Control, II, 1, 66-73.
KITAGAKI, I. (1996), Evaluation of students’ group using fuzzy integral, IEICE, J79-D-II, 11,
1888-1896.
KITAGAKI, I., SHIMIZU, Y. and SUETAKE, K. (1980), An instructional method which permits
the studentsto critically discuss their own test answers, Japan Journal of Educational
Technology, 5, 1, 23-33.
NAKATANI, A. (1995, in Japanese), Expert of interviewing, Diamond Co., Tokyo.



The 3
rd
International Conference on Virtual Learning, ICVL 2008

109
THE VIRTUAL TRAINING CENTRE (VTC) FOR CNC
(COMPUTER NUMERICAL CONTROL)

Catalin Dumitras
1
, Sahin Mehmet
2
, Mihai Aura
1
, Yaldiz Süleyman
2
,
Bilalis Nikolaos
3
, Maravelakis Emmanuel
4

(1) Gh Asachi Technical University of Iasi, Faculty of Leather and Textile Engineering, Romania
(2) Technical Science College, Selçuk University 42031, Konya, Turkey
mesahin@selcuk.edu.tr
(3) Department of Production Engineering & Management, Technical University of Crete,
73100, Chania, Greece
(4) Design & Manufacturing Laboratory, Technological Educational Institute of Crete,
73133 Chania, Greece


Abstract
The objective of this paper is to present the innovative training centre for CNC:
http://www.vtcforcnc.com. The Virtual Training Centre (VTC) was set up on the
Internet for Computer Numerical Control (CNC) training based on virtual aids. A
virtual space (a CNC training portal) on the Internet which allows the constant
sharing of e-learning-based CNC teaching material was created so as to foster the
further development of e-learning based CNC educational contents. The VTC for
CNC is an interactive platform, a meeting point for policy-makers, social-partners,
practitioners, researchers and all those with an interest in CNC field of vocational
education and training. Experts in the field are able to share and exchange knowledge
and experience with associates within and outside the European Union.

Keywords: Virtual Training, CNC, Virtual Environment


1. Introduction

Recently virtual training has been regarded as an innovation notably for vocational
training. There have appeared numerous virtual learning environments and various
approaches and tools to this end. The focus of “virtual learning” is in fact is on computer
technology and education. In this context, a large number of vocational training centres
and technical universities are giving priority to Computer Numerical Control (CNC)
Training, especially in the last decades. New developments on CNC machines are
providing a continuous need for updated CNC training curriculum. Training on CNC
should follow similar developments and in particular in their programming capabilities,
automation they offer and their technical capabilities. In addition, CNC programming is
becoming more and more automated through the use of CAD/CAM systems. This
requires from the programmers to acquire CAD operation capabilities, on top of their
CNC operation and programming knowledge. The major objective in the field of CNC
training is to improve the qualifications and competences of the trainees, which is directly
University of Bucharest and Ovidius University of Constanta


110
related to a well-designed and effective curriculum to be carried out on CNCs. The
facilities for CNC training vary a lot and this has had direct impact on the experience that
the trainee is acquiring during his/her apprentice.
This paper presents the design, the development of an International Virtual Curriculum
for CNC training, via an Internet based e-learning centre. This Virtual Curriculum in
CNC training is the main result of an International Leonardo da Vinci Project with three
participating European Countries. The developed training material is implemented in a
Virtual Training Centre (VTC), which includes a virtual space (a CNC training portal) on
the Internet which allows the constant sharing of e-learning based CNC teaching material,
and the further development of e-learning based CNC educational contents.
In order to develop the appropriate virtual content, the equipment, methods,
curriculum and techniques currently used for CNC training by the organisations in the
partner countries were observed, collected and evaluated. The selected materials were
used to create a new and common international curriculum. Five important factors that
contribute to learning were taken into account in order to prepare the CNC curriculum:
Motivation, Aptitude, Presentation, Repetition, and Practice with reinforcement. The
approach for developing the appropriate training material was based on these factors,
combined with carefully selected key concepts in CNC training. The result is a 28 session
Curriculum, implemented in the Virtual Training Centre, which aims at setting the
standard CNC virtual learning in vocational training systems.


2. CNC Training

Computer Numerical Control refers to the use of a computer to control and monitor
the movement of a machine. The machine could be a milling machine, lathe, router,
welder, grinder, laser or waterjet cutter, sheet metal stamping machine, robot or many
other types of machines. A CNC training course should consist of the tuition of CNC
programming methods and their application on actual conditions of processes. Its main
task should be to make any trainee at any training level capable of handling and
programming CNC machine tools.
CNC training usually takes place under supervisory attendance that emphasizes the
technological character of the training object. Additional support of appropriate teaching
material such as media and methods (slide-shows, movies, multimedia, demonstration of
manufactured pieces, visits to machine shops is often used. Furthermore laboratory
exercises are necessary for the understanding of each topic of the subject, some taking
place under actual conditions and other on paper. This way, the trainee can easier
understand the CNC machine programming, its applications and he can face the technical
problems encountered during the manufacturing of the parts. A large amount of
programming exercises can help the trainee to understand the theory in a better way,
offering him the sense of the quantity of the skills that has to obtain and the difficulties
that he is going to encounter, according to the machined part geometry. To accomplish all
these objectives, the exercises included in the curriculum, should include data from real
working conditions, as much as possible.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

111
3. A common CNC Curriculum

Each European Country has a different curriculum in CNC training. During the first
stages of the project, the equipment, methods, curriculum and techniques currently used
for CNC training by the organisations in the partner countries were observed, collected
and evaluated [1-3]. The selected materials were used to create a new and common
curriculum. Five important factors that contribute to learning were taken into account in
order to prepare the a common CNC curriculum:
• Motivation
• Aptitude
• Presentation
• Repetition
• Practice with reinforcement
The approach for developing the appropriate training material was based on the
following key concepts:
• Know your machine (from a programmer’s viewpoint)
• Prepare to write programs
• Understand the motion types
• Know the compensation types
• Format your programs in a safe, convenient, and efficient manner
• Know the special features of programming
• Know your machine (from an operator’s viewpoint)
• Understand the three modes of operation
• Know the procedures related to operation
• You must be able to verify programs safely
This approach combined with the important learning factors finally led to a CNC
training curriculum including 28 sessions:
(1) Machine configuration
(2) Speeds and feeds
(3) Visualizing program execution
(4) Understanding program zero
(5) Measuring program zero
(6) Assigning program zero
(7) Flow of program processing
(8) Introduction to programming words
(9) Preparation for programming
(10) Types of motion
(11) Introduction to compensation
(12) Dimensional (wear) tool offsets
(13) Geometry offsets
(14) Tool nose radius compensation
(15) Program formatting
(16) The four kinds of program format
(17) Simple canned cycles
University of Bucharest and Ovidius University of Constanta


112
(18) Rough turning and boring multiple repetitive cycle
(19) More multiple repetitive cycles
(20) Threading multiple repetitive cycle
(21) Subprogramming techniques
(22) Control model differences
(23) Other special features of programming
(24) Control model differences
(25) Machine panel functions
(26) Three modes of operation
(27) The key operation procedures
(28) Verifying new programs safely


4. Adaptation of the Curriculum into the virtual trainning center

In the report “Studies in the context of the E-learning Initiative: Virtual Models of
European Universities” [4], a key concern was how virtual mobility is being supported in
European universities through ICT integration and e-learning. The study found that the
majority of universities face major challenges in promoting ICT integration. ICT strategy
is very important and those universities that have an ICT strategy are significantly ahead
in integration of ICT in administration and organisation and networking. Integration of
ICT and e-learning is politically important in the EU in terms of internationalisation and
globalisation of education, student demand and interest in increasing the quality of
education through ICT [5-8]. At the national level, integration of ICT should become a
key priority with national and regional institutions making a commitment to ITC and the
development of networks. There must be increased national flexibility with a
commitment to support common standards of quality and assessment and to develop
national and international metadata standards. For all these reasons the designed common
curriculum for CNC was implemented into a Virtual Training Centre. To develop the
virtual training centre, a communication website was developed in order to manage the
activities and tasks to be carried out by the partners. Then, an interactive teaching
program was developed and put into a website to form a virtual training centre (figure 1).
The common curriculum developed for this purpose was the core of this training
centre. The site, along with the interactive teaching program, was divided into four main
areas, "News", "Exchange of views", "Projects and Networks", and "Information
Resources". With these, users would be able to access a newsletter, a bulletin board,
online surveys and survey reports, information on VET networks, an electronic library
with references, a bookshop with downloadable publications and a number of databases.
In the main core of the CNC training material, simulations and practical exercises are
included into the interactive training centre (see figure 2, 3, 4, ..).
The feedback of the implementation of the VTC in training centres has been
recorded and evaluated in order to produce the final version. The evaluation procedure
included content (topics, language used, modules), methods (progress, different levels of
difficulty, and range of resources, situations and practical cases) and technology (ease of
installation, interactive nature and use without a tutor).
The 3
rd
International Conference on Virtual Learning, ICVL 2008

113



Figure 1. Interface for http://www.vtcforcnc.com.

The main aim of the VTC for CNC aims is to be an interactive platform, a meeting
point for policy-makers, social-partners, practitioners, researchers and all those with an
interest in CNC field of vocational education and training. Experts in the field are able to
share and exchange knowledge and experience with associates within and outside the
European Union. This will foster the long-term viability of the Centre.



Figure 2. Interface for commands
University of Bucharest and Ovidius University of Constanta


114



Figure 3. An animation for M00 command



Figure 4. An animation on command M05
The 3
rd
International Conference on Virtual Learning, ICVL 2008

115



Figure 5. Interface for user simulation


5. Conclusion

The integration of ITC in this virtual learning environment for CNC, the
development of the VTC and the common training curriculum are focused on the EU
goals of internationalisation and globalisation of education, student demand and interest
in increasing the quality of education through ICT. At the national level, integration of
ICT has been a key priority with national and regional institutions making a commitment
to ITC and the development of networks. Furthermore, major national objectives include
an increased national flexibility with a commitment to support common standards of
quality and assessment and to develop national and international metadata standards. This
common Virtual CNC Curriculum addresses the priorities expressed here. Furthermore,
the Virtual Training Centre addresses the strategic objectives mentioned above:
improving the quality and effectiveness of education and training systems in the EU by
developing skills for the knowledge society, ensuring access to ICT for everyone,
increasing recruitment to scientific and technical studies, and making the best use of
resources. Facilitating the access of all to education and training systems by providing
open learning environment, making learning more attractive, and supporting active
citizenship, equal opportunities and social cohesion is the other strategic objective that
can be achieved through this virtual training centre. The experiences and knowledge
gained during the implementation of this Centre can be used in developing and improving
other training programmes in particular in the area of new information technology
applications in related sectors.
University of Bucharest and Ovidius University of Constanta


116
REFERENCES


ŞAHIN, M., BILALIS, N., YALDIZ, S., ANTONIADIS, A., ÜNSAÇAR, F., MARAVELAKIS, E., (2007),
Revisiting CNC Training–a Virtual Training Centre for CNC, International Conference on
E-Portfolio Process in Vocational Education-EPVET, Bucharest, Romania.
YADONG, LIUA, XINGUI, GUOA, WEI, LIA, KAZUO, YAMAZAKIA, KEIZO,
KASHIHARAB and MAKOTO, FUJISHIMAB, (2007), An intelligent NC program
processor for CNC system of machine tool. Robotics and Computer-Integrated
Manufacturing, Vol 23 (2), pp 160-169.
XIAOLING, W., PENG, Z., ZHIFANG, W., YAN, S., BIN, L., YANGCHUN, L., (2004),
Development an interactive VR training for CNC machining, Proceedings VRCAI 2004 – ACM
SIGGRAPH International Conference on Virtual Reality Continuum and its Applications in
Industry, pp. 131-133.
RAMBOLL, PLS, (2004), Studies in the context of the E-learning Initiative: Virtual Models of
European Universities (Lot1). Draft Final Report to the European Commission, DG
Education and Culture. Available at http://elearningeuropa.info
LARSON, J., and CHENG, H. H., Object-Oriented Cam Design through the Internet, Journal of
Intelligent Manufacturing, Vol.11, No. 6, December 2000, pp. 515-534.
ANDREATOS A., (2007), Virtual Communities and their Importance for Informal Learning,
International Journal of Computers, Communications & Control, Vol. II, No. 1, pp. 39-47
ZHENGXIN CHEN, Learning about Learners: System Learning in Virtual Learning Environment,
International Journal of Computers, Communications & Control, Vol. I (2008), No. 1.
K. SHEPPARD, G. KORFIATIS, S. MANOOCHEHRI, J. NASTASI, K. POCHIRAJU, E. MCGRATH,
P. DOMINICK, J. ARONSON (2004), Preparing Engineering Students for the International
Virtual Workplace : Information Technology Based Higher Education and Training, IEEE
Conference: ITHET 2004 pp 365- 370.





The 3
rd
International Conference on Virtual Learning, ICVL 2008

117
THE VIRTUAL TRAINING CENTRE FOR SHOE DESIGN
(VTC-SHOE): A MULTILATERAL VIRTUAL TRAINING
MODEL BASED ON A COMMON CURRICULUM

Aura Mihai
1
, Mehmet Sahin
2
, Süleyman Yaldiz
2
, Alina Dragomir
1
,
Nikolaos Bilalis
3
, Emmanuel Maravelakis
4

(1) Gh Asachi Technical University of Iasi,
Faculty of Leather and Textile Engineering, Romania
(2) Technical Science College, Selçuk University 42031, Konya, Turkey
mesahin@selcuk.edu.tr
(3) Department of Production Engineering & Management, Technical
University of Crete, 73100, Chania, Greece
(4) Design & Manufacturing Laboratory, Technological Educational Institute of Crete
73133 Chania, Greece


Abstract
Recently virtual training has been regarded as an innovation notably for vocational
training. There have appeared numerous virtual learning environments and various
approaches and tools to this end. The focus of “virtual learning” is in fact is on
computer technology and education. The objective of this paper is to present a virtual
training environment designed for shoe design training in the framework of EU LLP
projects: Virtual Training Centre for Shoe Design (VTC-SHOE). The aim of the
project is to implement shoe design training content (at elementary and intermediate
level) into a virtually designed and served training centre, which is accessible over
Internet, e-learning will be realised as an innovation in this field. Virtual Training
Centre for Shoe Design will be set up on the Internet to supply training (at
elementary and intermediate level) for shoe design. A virtual space, a shoe design
training portal on the Internet which will allow the constant sharing of e-learning
based shoe design training material so as to foster the further development of e-learning
based shoe design educational contents will be created. The equipment, methods,
curriculum and techniques currently used in shoe design training by partners will be
observed, collected and evaluated. The selected materials will be used to create a
new and efficient curriculum. This curriculum will be the core of target virtual
training activity to form the curriculum (at elementary and intermediate level).
According to this curriculum, an interactive teaching program will be developed and
put into a website to form a virtual training centre.

Keywords: Virtual Training, Shoe Design, Virtual Environment


1. Introduction

It is an accepted fact that the changing needs in training, in terms of both quantity
and quality, calls for promoting competitiveness and employment on the European
University of Bucharest and Ovidius University of Constanta


118
footwear industry. In order to foster use of information and communications technologies
(Commission Staff Working Document, 2001) in footwear industry, the Lifelong learning
programme can be regarded as an opportunity to overcome the challenges in this field by focusing
on the development of innovation and good practice (Decision No 1720/2006/Ec). One of
the major problems of the shoe industry at the moment is that the overall level of skills
and qualifications needs to be raised and, therefore, it is also necessary for training
modules to respond to the continuous evolution in the workplace so as to confront the
problem of unemployment and increased competition. Although the industry is asking for
shoe designers as professionals, the lack of training in this area, on various levels, has
been detected in some countries. A research by Mihai and Şahin (2007) displays that, for
example, according to the Romanian Occupational Standard, the “shoe designing” as a job or
its equivalent doesn’t exist as an occupation in VET. The curriculum for “Pattern Making
with Leather Products” for secondary level VET dates from 1998. The university level
course “Footwear Design and Technology” in Romania is too general and does not offer
all required competences essential for a Shoe Designer in line with the expectations in EU.
The main and comprehensive source that displays the situation of the footwear
sectors of the EU is the document titled Commission Staff Working Document: on the
promotion of competitiveness and employment on the European footwear industry. In this
document, it is strongly emphasised (on page 18) that the objective for the next ten years
is for "Europe to become the most competitive and dynamic knowledge-based economy
in the world". To achieve this the Commission has drawn up an action plan known as
eEurope, aimed at speeding Europe's transition to the information society and ensuring
that all Europeans possess the skills required for using the new information technologies.
Another report, titled Economic And Competitiveness Analysis Of The Footwear Sector In
The Eu 25, sets up that “training of human resources is also a way of investing in the
sector by helping workers to adapt to technological changes and to better face crisis situations”.
As stated by Mara Brugia (Helsinki, 20-23 June), “SMEs in Europe account for
99% of all businesses, and they provide employment for 74 million people. Decisive
factors of influence are: lack of a training culture within SMEs; lack of appropriate
training materials”. It can be inferred from this fact that almost every country in EU has
its own training materials, in some cases insufficient, and methods for shoe design
training. This brings about problems regarding the unification of workforce.
According to the report of Hanzl D., a.et. (2001), Romania had in 1999 about 66.000
employees in footwear industry, who produced 0,250 billion of EUR. The more recently
estimation of Eurostat (September, 2005) noticed for Romania footwear industry the following
data: number of enterprises-1400, turnover – 0,6 billion of EUR, value added at factor
cost – 0,2 billion of EUR, personnel cost – 0,2 billion of EUR, number of personnel
employed – 109100. It seams to be a high employment adds to, and in the same time, a
substantial reserve within possibility for future employment cut due with the increase the
productivity. Under such perspective, it is absolutely necessary for the actual and future
employees to take often up to date training courses in order to be competitive and well
fitted to the higher demands of the Romanian labour market in footwear sector.
According to the data by TASEV (Đstanbul, 2006), a foundation for research and
development of shoe making sector based in Istanbul, Turkey, the estimated number of the
graduates from the high schools and apprentice centres is about 1000 so far. The number of the
The 3
rd
International Conference on Virtual Learning, ICVL 2008

119
students attending Skill Acquisition Centres, High Schools and apprentice centres is about
720 now. The minimum expenditure for shoe design training in 2006 is about 1.800.000 USD. The
number of the firms in this sector is about 30.000 and the number of the employees is 300.000.
The rate of shoe export in Turkey in 1980s was not more than 2-3 million USD. In
1990s, the rate was about 30 million USD. The main boost was in 1990-2000, when it
was over 217 million USD. Present capacity is about 500 million pairs. A capacity of
about 200-300 million pairs is in use and the rest is not used. This unused capacity must
be activated. Here training is a very important problem. There are hardly enough qualified
staff who knows a foreign language. There are attempts to set up a training organisation
at university level. The first Shoe design department was opened at Technical Science
College of Selcuk University in 2006 and now there are 30 students.


2. Vocational education and training (VET) and ict use

The European Centre for the Development of Vocational Training (Cedefop) is the
European Union's reference Centre for vocational education and training. This centre
provides information on and analyses of vocational education and training systems,
policies, research and practice. According to Erwin Seyfried (2007), in the past two
decades and in most Member States there has been a growing awareness of the
importance of quality in vocational education and training (VET). Obviously, the
changing demands of the knowledge-based society and the overall trend to increase the
efficiency and effectiveness of VET systems, constitute major driving forces behind these
developments. Undeniably, through its funds and programmes, such as Leonardo da
Vinci, the European Commission has contributed to improving education and VET
systems by raising the level of the services they offer. For a qualitative approach to VET,
the technical working group on quality in VET (TWG) was called to respond to during its
mandate (2003 and 2004) in accordance with the priorities of the Council resolution of 19
December 2002 (Council Resolution of 19 December, 2002) and the Copenhagen
declaration on ‘enhanced cooperation in vocational education and training’ (European
Commission-DG EAC, 2004). Finally, a further focus of the work consisted of translating
the three European policy priorities (promoting employability of the workforce, access to
training with particular emphasis on the most vulnerable groups, and the better matching
of training demand and supply) into concrete and measurable objectives (Sahin at al., 2007).
One of the objectives of the innovative VET systems is regarded as transparency
and distribution of information. This function concerns the potential and actual use of
information. There may be different systems and structures of information distribution
among the various actors, and in the public. And there are preconditions for creating
transparency in the VET system. To improve quality there must be systems for
distributing information and certain mechanisms to ensure the circulated information can
be used by the various actors in the policy process. The more widespread the distribution,
the better the potential use of the data will be – and as a reversal effect, better quality data
can be expected, as the actors are able to check the information against their experience
and will provide feedback to the systems for gathering data.
University of Bucharest and Ovidius University of Constanta


120
One of the concrete future strategic objectives of education and training systems in
the EU (Council of the European Union, 2001) is improving the quality and effectiveness
of education and training systems in the EU. This includes improving education and
training for teachers and trainers, developing skills for the knowledge society, ensuring
access to ICT for everyone, increasing recruitment to scientific and technical studies, and
making the best use of resources. The second strategic objective is facilitating the access
of all to education and training systems. This objective includes open learning
environment, making learning more attractive, and supporting active citizenship, equal
opportunities and social cohesion.


3. Importance of virtual training in VET

During the 60's and 70's, teaching and learning tools were nothing but a piece of
chalk and a blackboard eraser, teachers and students who met each other face to face
inside the classroom during class. In the 80's, videotape programs were used as teaching
aids. In the 90's, one-way teaching by computer arrived. And finally today's advanced
computer and information network technology has revolutionized our teaching and
learning methods. In accord with the development, learning environment has also
changed. Students can listen to their teacher or trainers in distant classrooms through PC's
and get a simultaneous view of their teachers and texts as well. They can ask questions
and record the "class" for repeated viewing. Training organizations can conduct professional
training directly via the computer network. These learning environments are not so different
from a teacher-guided class with discussions and tests as well (Sahin at al., 2007).
In the report “Studies in the context of the E-learning Initiative: Virtual Models of
European Universities” (Ramboll, 2004), a key concern was how virtual mobility is being
supported in European universities through ICT integration and e-learning. The study
found that the majority of universities face major challenges in promoting ICT
integration. ICT strategy is very important and those universities that have an ICT
strategy are significantly ahead in integration of ICT in administration and organisation
and networking. Integration of ICT and e-learning is politically important in the EU in
terms of internationalisation and globalisation of education, student demand and interest
in increasing the quality of education through ICT. At the national level, integration of
ICT should become a key priority with national and regional institutions making a
commitment to ITC and the development of networks. There must be increased national
flexibility with a commitment to support common standards of quality and assessment
and to develop national and international metadata standards.


4. The aim of the paper

This article aims to promote a LdV project (Project Title: Virtual Training Centre
for Shoe Design, Project no: 134124-LLP-1-2007-1-RO-LEONARDO-LMP). This
project will address the strategic objectives mentioned above: The first one is improving
The 3
rd
International Conference on Virtual Learning, ICVL 2008

121
the quality and effectiveness of education and training systems in the EU by developing
skills for the knowledge society, ensuring access to ICT for everyone, increasing
recruitment to scientific and technical studies, and making the best use of resources. The
second one is facilitating the access of all to education and training systems by providing
open learning environment, making learning more attractive, and supporting active
citizenship, equal opportunities and social cohesion.



Figure 1. Interface of communication website (http://www.virtual-shoedesign.com)


5. The virtual training centre (VTC) for shoe design

The main objective of the project is to contribute to the development of quality
lifelong learning and to promote high performance, innovation and a European dimension
in system and practice in the field. VTC-Shoe project intends to improve vocational and
educational training curricula on shoe design in Romania, Turkey and Greece by focusing
on the development of innovation and good practice. The results of partners common
developments will be transpose into a virtual centre, making it available on European
level. By accessing the new created shoe design training course, trainers and teachers,
shoe designers, adult learners, as well as trainees and apprentice will be keeping up to
date with skills and knowledge necessary for high performance and innovation, both in
training and shoe design. Based on availability into virtual common space of the
innovative e-learning materials and training methodologies training materials, the project
will make its contribution to development of single European information space.
The second objective is to help promote creativity, competitiveness, employability
and the growth of an entrepreneurial spirit. In a world increasingly based on knowledge
and information, education and training are put at the core of the European footwear
industry agenda (Commission Staff Working Document, 2001). The footwear companies
need to make learning a lifelong endeavour deal with their employees of all ages
University of Bucharest and Ovidius University of Constanta


122
continuously developing their skills. By creating a new e-learning content and functional
web service the Virtual Training Centre for Shoe Design will help both workers and footwear
companies transforming the way they learn, interact and work in order to meet the footwear
sector needs for competitiveness, employability and the growth of an entrepreneurial spirit.
The third objective is to support the development of innovative ICT-based content,
services, pedagogies and practice for lifelong learning. ICT-related skills in the shoe
design are also vital for the competitiveness of the footwear sector from and for increased
job opportunities and employment. The concrete aim of the project is to develop a modern
virtual training centre in shoe design for: 1) training the trainers, trainees at the college
and technicians and apprentices for shoe design, 2) preparing shoe design technicians as
intermediates having common measurable qualities the industry is seeking. VTC-Shoe
project will create a common ICT-based content and will help for upgrading competences
and skills of teaching staff and exchange experiences over the virtual training centre.
As for the operational and specific objectives, the project aims to support
improvements in quality and innovation in vocational education and training system,
institution and practices. This can be achieved through improving the qualifications and
competences of the trainees in this field and it is directly related to the well-designed and
programmed curriculum to be carried out on shoe design. In addition, considering that
education is a dynamic process, it will be possible through this project, through its
dynamic and continuous characteristics, to improve the quality of vocational and
technical education, and accession to vocational training will be carried out.



Figure 2. Interface of the Virtual Training Centre (VTC) for Shoe Design
(http://www.vtcforshoedesign.com)


5.1. Methodological Approach

In order to strengthen and ensure that the project results will be used as regards the
target groups, target sectors and potential users, WP 2 (Developing Database for Team
Members of Target Sectors and Groups) will play an important role. This interactive
The 3
rd
International Conference on Virtual Learning, ICVL 2008

123
software for shoe design training based on new common curriculum will be served on
web. The log files will enable us to trace which of the expected users are logging on to
the site. The obtained results will be classified and compared to the expected usage levels
and the users will be encouraged and promoted to take part in dissemination.
The strategy of methodology and decision on the content development phases will
be determined basing on the common curriculum obtained in WP3. Thus, the stages to
form the content of VTC (Virtual Training Centre) need to: 1) consider the identified and
approved training needs/new qualification needs addressed by the common curriculum
designed in WP3, 2) consider the possible future developments. The content will be based
on the operability and comparability of the training program. Satisfaction of the target
groups will be of another concern. The responsible partners for this work package will
concentrate on the common curriculum produced in WP3. As well as that, the results drawn
up from the conference at the end of WP3 will be also a great asset for the content development
stage. As well as those, the partners will provide appropriate pedagogical experts from
their organisations to conduct and conclude the pedagogical approaches to the content.
This VTC-Shoe is a kind of library that contains a collection of training material
presented in many formats; documents, demonstrations, recorded lectures, and hands-on
labs. The VTC contents can be searched, reviewed all content of each type, or filtered the
content by a topic of interest. When one finds something that interests him or her, he/she
will click the link to access the content. The VTC-Shoe to be set up should contain six
stages under the following headings:
• The Introduction providing the setting.
• The Task telling the learner what to do.
• The Process suggests to the learner how to complete the task.
• The Resources are a set of website links or other resources like the common
curriculum (on or offline) that the learner will use to find the appropriate
information.
• The Evaluation informs the learner as to judge the success. This could also
include soft outcomes.
• The Conclusion rounding up the activity.


5.2. Innovative Aspects

Virtual Training Centre for Shoe Design will be set up on the Internet to supply
training for shoe design. A virtual space, a shoe design training portal on the Internet
which will allow the constant sharing of e-learning based shoe design training material so
as to foster the further development of e-learning based shoe design educational contents
will be created. The equipment, methods, curriculum and techniques currently used in
shoe design training by partners will be observed, collected and evaluated. The selected
materials will be used to create a new and efficient curriculum. This curriculum will be
the core of target virtual training activity. Therefore, four main aspects will be solved:
1. Virtual training will be an innovative approach in shoe design training, as the Virtual
Training Centre will include interactive training software based on the new and common
curriculum. In this way more trainees will benefit from the same and reliable source of training.
University of Bucharest and Ovidius University of Constanta


124
2. The Simulation program served on Virtual Training Centre will be expected to solve
lack of unification of the shoe design tools necessary for training purpose for numerous trainees.
3. The common curriculum will enable the partner countries to cooperate in shoe
design training. Thus the authorities will be able to obtain mutually a means to measure
the level of outcomes in shoe design training in partner countries.
4. Creating an interactive teaching program served on Internet will help even while at
work and disseminating information regarding the latest technology and innovations
through observation and implementation will be beneficial to the shoe manufacturers.
The general features of the anticipated web platform will be: easily comprehensible
(readability of the codes will be quite high), web modularity (a modular structure that will
be accessed by different user groups (a modular structure that is provided by object
oriented programming techniques), separation of supplying and operating logics (the
system will conduct operational functions and the design of user interface separately from
the operation logics). By means of this understanding, a secure structure will be obtained
basing on the web manager and user modules.
a) Web Manager will be in an Internet application structure that will provide the
administration of the web site and provide the web broadcasting of content formers over
web browsers.
• Technical Features: MySQl will be used as database. For the confirmation of the
registered users, a special index keeping the information of the users will be used.
• Development of application process: P1 will be responsible for the coordination
of the development of the applications. Algorithmic infrastructure, software
development, hosting procedures, creation of the web platform and necessary
development procedures will be done by one or two sub-contractor. Development
of the security module, make use of the different databases and testing the system
at different stages will be under the responsibility of P1.
• Basis of the program: This application will be developed enabling the distance
web site managing. Thus, the construction and update of partner sub-web sites
both in English and in native languages will be possible by using the English
based management modules. The partners will transfer the information for their
sub-web site (under the original web site) to the web server over this application.


5.3. European Added Value

The rapidly changing technologies, as well as the innovative e-learning teaching methods
require for adapted modules for lifelong training that keeps continuously up to date with
the relevant developments of the European footwear industry. The Virtual Training Centre
for Shoe Design will be an interactive platform, a meeting point for policy-makers,
social-partners, practitioners, researchers and all those with an interest in shoe design
field of vocational education and training. Experts in the field will be able to share and
exchange knowledge and experience with associates within and outside the European Union.
The project’s scientific and pedagogic objectives are in tune with the main priority
in Lifelong Learning Programme: Part I (EAC/61/2006). Through the various research
The 3
rd
International Conference on Virtual Learning, ICVL 2008

125
and development projects partner P1 has developed training materials for shoe design. These
materials have to be compared between involved partners in order to get common curricula
to be share with future users at a European level. The innovative e-content, which will be
developed within the VTC-Shoe project, can easily be translated to various languages.
In terms of Strategic Impact and Contribution to growth, the VTC-Shoe project is
expected to have a very powerful impact in the European Footwear industry. Closely to
the other projects funded by European Community, it will improve competitiveness
helping footwear companies to have skilled and competent shoe designers.
Thus, VTC-Shoe added value for the Community lies in the provision of a training
tool that has the dynamics not only to provide valuable training and skills to the targeted
beneficiaries but also to empower the processes of the EU Clothing Industry and thus,
increase productivity and competitiveness. This, in its turn, is expected help the industry
grow and, thus, increase the demand for more skilled employees.
This virtual training centre to be formed in this field and its application will set the
first and good example for virtual learning in national vocational training systems. It will
help improve and upgrade competences and skills of staff and exchange experiences over
the virtual training centre. It will increase the work opportunity by helping young
generation to use Information Technologies.
As this project contributes to e-learning by providing new training tools, it will
create new job opportunities for the individuals in partner countries and thus this will
contribute to employment exchange in EU. These activities are inline with European
strategies for vocational training. As emphasised at LEARNTEC 2005 (Karlsruher
Messe- und Kongress-GmbH, Tuesday 02nd of November 2004), “E-Learning has
become indispensable for corporate training, but the fascination of SMEs for web-based
learning is still quite limited. Further education and training figures rarely on the priority
list, and in most of the cases, a training department does not even exist.” This project is
also expected to help the participants to acquire required qualities, to be proud of this, and
in this way, to be an active citizen of EU.
The internet based platform within the VTC-Shoe project offers to trainers/teachers
the possibility for continuing development of their skills and competences. The innovative
solutions for training in shoe design as well as the innovative pedagogical methodologies
will keep them up to date with the new technologies in order to have a longer active
professional life. Within the shoe industry, the majority of workers are women (Nina Ascoly &
Chantal Finney, 2005) and account for about 40% to 50% of all employees. Our project
will open new training opportunities for women. However, the e-learning shoe design course
developed within the Virtual Training Centre is an equal opportunities course, combating
all forms of discrimination based on sex and will be open to both man and women
One of the major problems of the footwear industry at the moment is that the
overall level of skills and qualifications needs to be raised and, therefore, it is also
necessary for training modules to respond to the continuous evolution in the workplace so
as to confront the problem of unemployment and increased competition. VTC-Shoe
project proposal comes up to the changing needs in training, in terms of both quantity and
quality, designed for promoting employment on the footwear industry. Training materials
offered by the VTC-Shoe internet platform will help both unemployed people to find a
job in footwear companies, and worker to up to date their skill for getting a better
University of Bucharest and Ovidius University of Constanta


126
position. The Virtual Training Centre (VTC) will be set up on the internet for Shoe
Design training. This will allow the constant sharing of e-learning based Shoe Design
teaching material will be created so as to foster the further development of e-learning
based Shoe Design educational contents. The VTC for Shoe Design will be an interactive
platform, a meeting point for policy-makers, social-partners, practitioners, researchers
and all those with an interest in Shoe Design field of vocational education and training.
Experts in the field will be able to share and exchange knowledge and experience with
associates within and outside the European Union.
The free toolkit, available on this site, will provide the trainers and trainees with the
facility to design the product in mind. Access to create, update or modify the VTC-Shoe will
be available throughout the project and after it. The interactive teaching program served
on internet will help even while at work and disseminating information regarding the
latest technology and innovations through observation and implementation will be
beneficial to the manufacturers. This interactive software for Shoe Design training based
on new common curriculum will be served on web. The log files will enable us to trace
which of the expected users are logging on to the site.


6. Conclusion

This virtual learning environment for Shoe Design as an e-learning environment is
important in the EU in terms of internationalisation and globalisation of education,
student demand and interest in increasing the quality of education through ICT. At the
national level, integration of ICT should become a key priority with national and regional
institutions making a commitment to ITC and the development of networks. There must
be increased national flexibility with a commitment to support common standards of
quality and assessment and to develop national and international metadata standards. This
centre addresses the priorities expressed here. Furthermore, this virtual training centre addresses
the strategic objectives mentioned above: improving the quality and effectiveness of
education and training systems in the EU by developing skills for the knowledge society,
ensuring access to ICT for everyone, increasing recruitment to scientific and technical
studies, and making the best use of resources. Facilitating the access of all to education
and training systems by providing open learning environment, making learning more
attractive, and supporting active citizenship, equal opportunities and social cohesion is the
other strategic objective that can be achieved through this virtual training centre.
On a short term, the partners country will have trainers from colleges, vocational
schools being up to date with the a new common curricula and having necessary skill for
teaching on-line; trainees with more extensive knowledge in shoe design, more skilled
design technicians, designers, who are actually responsible for designing shoes and
apprentices [prepared for a new job. who are newly recruited for shoe design.
The Virtual Training Centre for Shoe Design is necessary for Universities, footwear
companies, colleges and training institutions all over Europe and elsewhere, because they
are integrating in an organised and illustrative way all the steps required to acquire
quickly, easily and in a technologically advanced manner the skills necessary for shoe
The 3
rd
International Conference on Virtual Learning, ICVL 2008

127
design, and pattern construction and which will be more clearly and in a more effective
educational approach than in an ordinary classroom. Through the network of collaborations of
the partnership, the outputs of the training tools will be assimilated in the training systems
of a wider spectrum of training organisations.
The results of the project can be transferred to similar fields such as furniture,
textile, air conditioning etc. The experiences and knowledge gained during the process of
this project can be used in developing and improving other training programmes in
particular in the area of new information technology applications in related sectors. The
final form of the training programme will be publicised on the co-ordinating
organisations’ website. In addition, through meetings, workshops and conferences and
seminars to be held with the related institutions, the results will be introduced to other
sectors. As this type of learning is virtual and based on distance learning, it will be
possible to have access in all geographical contexts.



REFERENCES


Commission Staff Working Document: on the promotion of competitiveness and employment on the
European footwear industry. This report was issued in Brussels, on 28.2.2001, SEC (2001) 366.
Council of the European Union, 2001.
Council Resolution of 19 December 2002, 2003.
Decision No 1720/2006/Ec of The European Parliament and of the Council Of 15 November 2006,
Art.25, pct. b, art. 26, pct. 1.d.
Decision of the European Parliament And Of The Council establishing a Competitiveness and
Innovation Framework Programme (2007-2013), COM (2005) 121 final, 2005/0050 (COD).
Economic and competitiveness analysis of the footwear sector in the EU 25”, September 2005.
European Commission – DG EAC, 2004.
HANZL, D., BRETON, P., and JANUSKAITE, R., Competitiveness of Industry in Candidate
Countries, 2001, under the framework of contract PSE/99/502333.
Lifelong Learning Programme: Part I – Priorities Of The 2007 General Call For Proposals, (Eac/61/2006).
MARA BRUGIA, in Eden 2005 Annual Conference, Helsinki, 20-23 June.
MIHAI AURA, MEHMET ŞAHIN, E-Portfolio in Vet: A Study into the Link between Personal
Development Planning and Curriculum in Shoe Design Training, EPVET 2007: International
Conference on E-Portfolio Process in Vocational Education, Present and Future, 2-3 May 2007,
Bucharest, Romania.
NINA ASCOLY & CHANTAL FINNEY (editors), 2005, Made by Women, Gender, The Global
Garment Industry And The Movement For Women Workers’ Rights, http://www.cleanclothes.org/
ftp/made_by_women.pdf, downloaded on 28.03.2007.
Ramboll, PLS, (2004): Studies in the context of the E-learning Initiative: Virtual Models of
EUROPEAN Universities (Lot1). Draft Final Report to the European Commission, DG
Education and Culture, available at http://elearningeuropa.info
ŞAHIN, M., BILALIS, N., YALDIZ, S., ANTONIADIS, A., ÜNSAÇAR, F., MARAVELAKIS, E., 2007,
Revisiting CNC Training – A Virtual Training Centre for CNC, International Conference on
E-Portfolio Process in Vocational Education-EPVET, Bucharest, Romania.
University of Bucharest and Ovidius University of Constanta


128
ŞAHIN MEHMET, SÜLEYMAN YALDIZ, FARUK ÜNSAÇAR, Y. BURAK YALDIZ,
NIKOLAOS BILALIS, EMMANUEL MARAVELAKIS, ARISTOMENIS ANTONIADIS,
Virtual Training Centre For CNC: A Sample Virtual Training Environment, ICVL 2007:
The 2nd International Conference on Virtual Learning, 26-28 October, 2007, Constanta, Romania.
SEYFRIED ERWIN, Evaluation of Quality Aspects in Vocational Training Programmes: Synthesis
Report, CEDEFOP, 1998.
TASEV: Türkiye Ayakkabı Sektörü Araştırma, Geliştirme ve Eğitim Vakfı, Đstanbul, 2006.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

129
Virtual Learning Environments and World Languages
The Way Forward
The Flexi-Pack Project as SOAS-UCL CETL
(University of London)

Nathalie Ticheler

SOAS-UCL Centre for Excellence in Teaching and Learning Languages of the Wider World,
School of Oriental and African Studies Room 443
Thornaugh Square, London WC1H 0XG, UNITED KINGDOM
E-mail: nvticheler@yahoo.it


Abstract
The SOAS-UCL Centre of Excellence in Teaching and Learning ‘Languages of the
Wider World’ at the University of London aims to promote excellence in the
teaching and learning of languages that do not have a large presence in UK Higher
Education but which are of increasing strategic importance, locally and globally. A
key objective of the CETL is to support blended language learning. Flexi-Packs, one
of our flagship projects, is an innovative package of specially-tailored learning
materials, designed to promote mobile learning, with opportunities for learner
empowerment and collaborative learning. The paper will present the pedagogical
rationale behind the Flexi-Packs and reference will be made to Romanian materials
produced at University College London.

Keywords: Virtual Learning Environments, World Languages, mobile learning, learner
empowerment, collaborative learning.


1. Introduction

The precarious situation of Modern Foreign Languages (MFL) in the United
Kingdom (UK), with issues such as the decreasing number of students on specialist
language degree courses and the closure of university departments, is reported by
numerous organisations such as the Centre of Information for Language Teaching (CILT)
and the Higher Education Funding Council for England (HEFCE), as well as in the
Nuffield Language Inquiry Reports (2003).

According to official figures from the Higher Education Statistics Agency, around 71,000
students were taking at least one accredited module in languages in 2001/2002. This represents
4.6% of all students and has declined from 90,000 (6.4%) in 1998/1999. (Kelly & Jones, 2003)

Considerable concern has been expressed in the press about the long-term future of
languages in UK schools and universities and about the implications for business. (CILT, 2005)

University of Bucharest and Ovidius University of Constanta


130
In this context, various initiatives have been implemented in an attempt to improve
the precarious situation of MFL in the UK. The National Languages Strategy, launched in
2002, has implications at all stages of the education system and extends beyond the
classroom, including at international level:

In the knowledge society of the 21
st
century, language competence and intercultural
understanding are not optional extras, they are an essential part of being a citizen. (Ashton, 2002)

In addition, HEFCE has agreed to fund a programme to encourage the take-up of
language courses in England. Routes into Languages was originally funded from
HEFCE's Strategic Development Fund. The programme is running for four years from
2006/07 to 2009/10. It will be led by the Subject Centre for Languages, Linguistics and
Area Studies (LLAS), in a partnership with the University Council of Modern Languages
(UCML) and the Centre for Language Teaching (CILT).
In a context of significant evolutions in Higher Education, such as the widening
participation of students from non-traditional social and educational backgrounds,
together with the necessity to operate within budgetary constraints, e-learning is
presented as the ideal answer to current requirements, both at students’ level and at
institutional level.

Increasing diversity in the student population, through widening participation, new
technologies and new, more cost-efficient practices in course production are forcing a re-think
of current activity and providing a challenge to all those involved in the design and delivery
of learning constantly seek out ways of ensuring that the needs of our language learners are
met.(Hurd, 2002)

The economic necessity of linking Information and Computer Technology with
education is perhaps the most prominent strand of the rhetoric surrounding learning
technologies in post-compulsory education. However, the majority of the rhetoric
surrounding learning technologies has centred on the individual learners, in particular
the empowerment of the individual’s learning experience:

Until now, learning has tended to be static and fixed. Learners have had to go to a
site of learning such as a college or school at specific times. E-learning can change all this.
Learners can choose what, how and when they learn and learning can now be defined by
those choices, rather than by the time available to attend a physical centre of learning.
(DFES, 2002)

In March 2005, the DFES presented a five-year e-learning strategy Harnessing
Technology: Transforming Learning and Children’s services, with implications in all
areas of education, from primary schools to universities. ICT is clearly shown as a
participational and motivational tool:

At any stage of learning, ICT could re-engage the unmotivated learner” (DFES,
2005) and “the new technologies are capable of creating real energy and excitement for all
age groups. Used well, they should motivate, personalise and stretch. (DFES, 2005)

The 3
rd
International Conference on Virtual Learning, ICVL 2008

131
The CETL Languages of the Wider World is a collaboration between University
College London and the School of Oriental and African Studies and is funded by the
Higher Education Funding Council for England. The CETL aims to promote, foster and
support excellence in the teaching and learning of languages of the Wider World. It
covers four main themes: reflection and research, materials and curriculum development,
learner and teacher training, as well as dissemination.
The Flexi-Pack project, which focuses on m-learning, is based at and funded by the
SOAS-UCL Centre of Excellence for the Teaching and Learning Languages of the Wider
World. CETL launched a call for bids for Flexi-Pack funding in May 2007 and
applications were successful in the following languages: Romanian at University College
London, Bengali, Nepali, Japanese and Turkish at the School of Oriental and African
Studies. The project has recently been completed and is now being piloted with students.
It is now intended to increase funding in this area.


2. Mobile Learning

In recent years, mobile learning (m-learning) has attracted a great deal of interest
throughout the educational spheres and this has resulted in numerous pilot projects and
research papers. An example of this is the review conducted by Cobcroft et al (2006)
which concerned over 400 publications on m-learning including conference papers,
reports, reviews and research projects.
M-learning may currently be considered as a loosely-defined concept, with a wealth
of initiatives related to the use of handheld devices both in and out of the classroom,
including for self-study and to supplement taught sessions. There is a need to clarify the
definition and scope of m-learning and, in this paper, m-learning is described as the use of
handheld devices outside the classroom by adult students for learning purposes in Higher
Education contexts in the United Kingdom.
M-learning concerns the acquisition of knowledge and skills through the use of
mobile technology, irrespective of time and location. (Geddes, 2004)

the term “mobile learning” is frequently used to refer to the use of handheld technologies
enabling the learner to be on the move, providing anytime anywhere access to learning.
(Price, 2007)

M-learning gives us the opportunity to design learning differently, to create extended
learning communities, to provide expertise on demand, and to support a lifetime of learning.

mobile learning is not just about learning using portable devices, but learning
across contexts. (Sharples, 2007)

Researchers such as Attewell present the advantages of m-learning, which concentrate
primarily around personalisation of learning, collaborative learning, a greater informality
of the learning experience and an enhanced engagement of reluctant learners.

University of Bucharest and Ovidius University of Constanta


132
Mobile learning is unique in that it allows truly anywhere, anytime, personalised
learning. It can be used to enrich, enliven or add variety in conventional lessons or
courses.” and also “Mobile learning helps to remove some of the formality from the
learning experience and engages reluctant students. (Attewell, 2005)

Kukulska-Hulme, another supporter of m-learning, insists that learning technologies
have ceased to be the preserve of technicians and experts and that teachers and learners
have begun to integrate them into their normal daily practice. For her,

Mobile learning promises to deliver closer integration of language learning with
everyday communication needs and cultural experiences. (Kukulska-Hulme, 2006)


3. Snapshot on the Romanian Flexi-Packs

Romanian Flexi-Packs produced at University College London, with funding from
the Centre of Excellence in Teaching and Learning Languages of the Wider World, are
specially-tailored materials designed for e-learning and m-learning, in accordance with
principles of blended learning. Their purpose is to supplement traditional taught sessions
with online materials tailored to students’ needs, both in terms of contents and level of difficulty.
Romanian Flexi-Packs are based on a Virtual Learning Environment and available
for learners to download and use “on the go” for mobile learning. Every week, students of
Romanian have the opportunity to download a text file in PDF format, together with a set
of mp3 audio-files. A typical text file normally contains sections such as: learning
objectives, learning tips, reference sections with vocabulary and grammar, a carefully-
selected list of web sites with additional tasks to complete, a whole range of exercises
which correspond to the contents of the previous lesson and cover listening, reading and
writing skills, as well as grammar and vocabulary and finally, transcripts for all the audio-
files and keys to all the activities to allow for students’ self-assessment. Flexi-Packs focus
on functional language and aim to provide students with real-life snapshots on Romania
in context, offering topics such as meeting people and travelling on the train.
Twenty Romanian Flexi-Packs, suitable both for undergraduates and post-graduates
have been produced at University College London. They contribute to reducing the lack
of suitable published materials identified by tutors, especially regarding listening and
speaking skills. Romanian Flexi-Packs have been piloted among students, who have
responded very positively to their new m-learning experience and have made the
following comments: “it really is what I was looking for”, “Flexi-Packs are
straightforward to use” and “with Flexi-Packs, it is easier to progress”.
Flexi-Packs correspond to the objectives of the HEFCE e-learning strategy launched in
2005, in particular regarding the diversity of learners’ needs and the flexibility of
provision. Another measure of success employed by HEFCE in its e-learning strategy is
tutors’communication with the students, as well tutors’ access to materials for regular use
and improvement. The Flexi-Pack Project has taken these points into consideration.
Flexi-Packs offer a variety of advantages, with a view to maximise students’
experience of m-learning. First of all, Flexi-Packs are specially-tailored to the students’
The 3
rd
International Conference on Virtual Learning, ICVL 2008

133
needs and give them a greater say in what they learn, when and how. Then, Flexi-Packs
are easy to modify by tutors, even with limited technical knowledge and this applies in
particular to text files. Moreover, they offer flexibility in terms of presentation, as they
can be placed on web pages or on Virtual Learning Environments for students to
download and use for m-learning.


4. Concluding Notes

This article outlines the rationale for creating Romanian Flexi-Packs, a range of m-
learning materials with a fully-integrated approach between traditional lessons and self-
study. Flexi-Packs are currently being developed for a variety of languages including
Turkish, Bengali, Nepali and Japanese at SOAS and UCL. Flexi-Packs truly offer a m-
learning experience to students, with materials which correspond to their needs and allow
for a greater empowerment in their learning experience. More opportunities for further
development are available through a greater use of Virtual Learning Environments. For
further information about Flexi-Packs, or about CETL work more generally, please visit
http://www.lww-cetl.ac.uk



REFERENCES


ALLY, M. (2004), Using Learning Theories to design Instruction for Mobile Learning
Devices. In M-Learn 2004 Conference Papers, Rome, 5-7.
ATTEWELL, J. & WEBSTER, T. (2004), Engaging and Supporting Mobile Learners.
In M-Learn 2004 Conference Papers, Rome, 15-19.
ATTEWELL, J. (2005), Mobile Technologies and Learning. A Technology Update and M-Learning
Project Summary. Learning and Skills Development Agency, London.
BEALE, R. (2004), Wireless Learning community Hub. In M-Learn 2004 Conference Papers,
Rome, 23-24.
BULL, S. et al. (2004), Interactive Logbook: The Development of an Application to Enhance and
Facilitate Collaborative Working within Groups in Higher Education, in M-Learn 2004
Conference Papers, Rome, 39-42.
COBCROFT, R. et al. (2006), Literature Review into Mobile Learning in the University Context.
Queensland University of Technology, Queensland.
CILT, (2005), Language Trends. CILT, London, http://www.cilt.org.uk/research/languagetrends/
2005/trends2005_community.pdf
CILT, (2007). Higher Education Statistics. Analysis of HESA data. CILT, London, http://www.cilt.
org.uk/research/statistics/education/higher.htm
CONOLE, G. (2004), E-Learning: The Hype and the Reality. Journal of Interactive media in Education.
DFES (2002), National Languages Strategy. DFES, London.
DFES (2005), Harnessing Technology. Transforming Learning and Children’s Services. Summary
Version (3
rd
report). DFES, London.
DORNYEI, Z. (1997), Psychological Processes in co-operative language learning group dynamics
and motivation. The Modern Language Journal 81, 482-493.
University of Bucharest and Ovidius University of Constanta


134
FERSHA, A. et al. (2004), Team Awareness in Personalised Learning Environments, in M-Learn
2004 Conference Papers, Rome, 69-72.
GEDDES, S. J. (2004), Mobile Learning in the 21
st
Century: Benefits for Learners. The Knowledge
Tree E-journal, 6.
HEFCE (2006), Routes into Languages. HEFCE, London, http://www.routesintolanguages.ac.uk
HURD, S. (2002), Learner Differences in Individual Learning Contexts, http://www.llas.ac.uk/
resources/goodpractice.aspx?resourceid=1573
KELLY, M. and JONES, D. (2003), A new Landscape for Languages in Nuffield Language Inquiry
Reports, The Nuffield Foundation, Southampton.
KUKULSKA-HULME, A. (2006), Mobile Language Learning Now and in the Future. The Open
University, Milton Keynes.
NAISMITH, L. (2006), Literature Review in Mobile Technologies and Learning. Futurelab, Birmingham.
ONUALLAIN, C. & BRENNAN, A. (2004), How can one effectively assess students working in a
collaborative mobile environment on an individual basis, in M-Learn 2004 Conference
Papers, Rome, 149-152.
PRICE, S. (2007), Ubiquitous Computing: Digital Augmentation and Learning, in N. Pachler
(ed.), Mobile Learning: Towards a Research Agenda. Occasional Papers in Work-Based
Learning 1. WLE Centre for Excellence, London.
SHARPLES, M. et al. (2005), Towards a Theory of Mobile Learning. Mlearn, Birmigham, http://www.mlearn.
org.za/CD/papers/Sharples-%20Theory%20of%20Mobile.pdf
SHARPLES, M. (2007), Foreword. Big Issues in Mobile Learning. Kaleidoscope, Nottingham.
TRAXLER, J. & BRIDGES, N. (2004), Mobile Learning: The Ethical and Legal Challenges, in
M-Learn 2004 Conference Papers. Rome, 203-207.
WALKER, K. (2007), Mapping the Landscape of Mobile Learning. Big Issues in Mobile Learning,
Kaleidoscope, Nottingham.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

135
Applying 0/1 Integer Programming to Optimize User Curriculum
in a Virtual Learning Environment based on Utility Function

Hamed Fazlollahtabar
1
*, Morteza Mofidi
2

1
Young Researchers Club, Islamic Azad University Babol Branch, Iran
2
Department of Industrial Engineering Mazandaran University of Science and Technology,
Babol, Iran
*E-mail: hamed@http.ac.ir



Abstract
The Internet and the World Wide Web in particular provide a unique platform to
connect learners with educational resources. Educational material in hypermedia
form in a Web-based educational system makes learning a task-driven process. It
motivates learners to explore alternative navigational paths through the domain knowledge
and from different resources around the globe. Consequently, many researchers have
focused on developing e-learning systems with personalized learning mechanisms to assist
on-line Web-based learning and to adaptively provide learning paths. However,
although most personalized systems consider learner preferences, interests and
browsing behaviors when providing personalized curriculum sequencing services,
these systems usually neglect to consider whether learner ability and the difficulty level of
the recommended curriculums are matched to each other. Therefore, our proposed
approach is based on Integer programming (IP) to optimize user curriculum based
on the utility function which is related to preferences.

Keywords: virtual learning environment (VLE); integer programming; user curriculum;
utility function.


1. Introduction

Virtual Learning (VL) system is an internet based service like the application
system or the internet based virtual course study service (this paper argues this service to
be a part of the e-Learning system). This system is able to be interpreted in various ways
such as ‘‘computer based, education delivery system which is provided through the
Internet’’, or ‘‘an educational method that is able to provide opportunities for the needed
people, at the right place, with the right contents, and the right time’’ (Song, 2000).
The e-Learning system is one of many methods of the education (the teaching and
learning procedure) that allows flexible learner-centered education. It is an information
system based on the World Wide Web. E-Learning provides an inter-disciplinary approach to
information technology and educational engineering, and an assessment of e-Learning
effectiveness could also be achieved. As of IT, the end user assessment, the quality of the
information system, and the system’s user satisfaction could be measured.
University of Bucharest and Ovidius University of Constanta


136
As of educational engineering, however, the learner’s academic achievement or the
degrees of self-study ability could be measured. The academic achievement is an
assessment of the learner’s e-Learning environment, while self-study ability is an
assessment of one’s aptitude regarding his or her self-study. This approach reveals the
extensive and effective trends resulting from an e-Learning research. Many researchers
are quite divided over the various views regarding educational engineering and
information systems. Many researchers are on exploratory level trying to get explanations
regarding the variations of e-Learning effectiveness (i.e., Wang, 2003).
The tendency of educational engineering to introduce theoretical variables in order
to explain e-Learning effectiveness is insufficient except for limited numbers of
information systems (i.e., Piccoli, Ahmad, & Ives, 2001). Moreover, this approach of
putting together information systems and educational engineering is rarely observed.
Integer Programming (IP) is an important technique for dealing with problems that
arise frequently in diverse fields such as capital budgeting, production planning, capacity
planning, scheduling and chemical engineering process. These applications are
extensively surveyed in Djerdjour (1997), Salkin and Mathur (1989), Simons (1996) and
Taha (2003). Biegler and Grossmann (2004) provided a retrospective article on
mathematical programming models and optimization techniques that have been applied in
process systems engineering. They indicated that operations problems give rise to mixed-
integer linear programming (MILP) and mixed-integer non-linear programming (MINLP)
models for scheduling and supply chain problems. Besides, the design and synthesis
problems and the hybrid systems in the control problems have also been formulated as
MILP and MINLP problems. Since the enterprise can achieve a cost advantage by
reducing costs from the current solution, optimization techniques have received extensive
attention from the practitioners and the researchers in the last few decades.
A number of applications require to model discrete variables, such as number of
units or batches, ‘‘yes or no’’ decision, sequences in scheduling, etc. Generally, a
problem with integer variables, zero-one variables and mixed variables are called IP
problems. Because of their importance in formulating many practical problems, there has
been a pronounced increase in the development of optimization approaches on IP models.
These approaches can mainly be classified into stochastic and deterministic.
In this paper we first, propose a curriculum network for users in a virtual learning
environment and then design a utility function to select the profile. Finally, the optimal
curriculum is chosen applying 0/1 integer programming approach.


2. The Proposed User Curriculum Network

In this paper, we propose a virtual learning environment network. The basis of that
network is that a student wants to choose his educational curriculum based on a standard.
The virtual learning system is including departments, courses, and teachers. A student
should choose a department, based on his learning profile, in which different courses are
offered. Each course is provided with some teachers. Those selections are based on a
standard. The details about different standards and the applied standard in this paper are
explained in the next section. The overall scheme of the proposed VLE is indicated in Figure 1.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

137

Student
Department 1 Department 2 Department m
...
Course 1 Course 2 Course n Course 1 Course 2 Course n Course 1 Course 2 Course n
...
Teacher 1 Teacher 2
Teacher s
ij


Figure 1. The Proposed VLE Network


3. Utility Function

A utility is a numerical rating assigned to every possible outcome a decision maker
may be faced with. (In a choice between several alternative prospects, the one with the
highest utility is always preferred.) To qualify as a true utility scale however, the rating
must be such that the utility of any uncertain prospect is equal to the expected value (the
mathematical expectation) of the utilities of all its possible outcomes (which could be
either "final" outcomes or uncertain prospects themselves). When decisions are made by a
so-called rational agent (if A is preferred to B and B to C, then A must be preferred to C),
it should be clear that some numerical scale can be devised to rate any possible outcome
"simply" by comparing and ranking these. Determining equivalence in money terms may
be helpful in such a systematic process but it's not theoretically indispensable. What may
be less clear, however, is how to devise such a rating system so that it would possess the
above fundamental property required of a utility scale.
One theoretical way to do so is to compare prospects and/or final outcomes to
tickets entitling the holder to a chance at winning some jackpot, which is at least as
valuable as any outcome under consideration. A ticket with a face value of 75% means a
University of Bucharest and Ovidius University of Constanta


138
chance of winning the jackpot with a probability of 0.75 and it will be assigned a utility
of 0.75. Anything which is estimated to be just as valuable as such a ticket (no more, no
less) will be assigned a utility of 0.75 as well.
The scale so defined does have the property required of utility scales. Consider, for
example, a prospect which may have one of two outcomes:


• The first outcome has a probability of 0.3 and a utility of 0.6 (it could be a ticket
with a 60% face value).
• The second outcome has a probability of 0.7 and a utility of 0.2 (it could be a
ticket with a 20% face value).


The prospect has therefore, by definition, a utility of 0.32, and we do observe that
the result has been computed with the same rule as a mathematical expectation. It would
be so in any other case involving either lottery tickets or things/situations previously
assigned a utility (by direct or indirect comparisons with such tickets).
The types of utilities introduced above are between 0 and 1, but no such restriction
is in fact required. The key observation is that we may either translate or rescale a utility
scale without affecting at all the decisions it implies: Each side of every comparison is
translated or rescaled the same way and it does not affect inequalities as long as the
scaling factor is positive. In particular, we may keep the same utility scale if we're faced
with an outcome more valuable than whatever jackpot we first considered. If that jackpot
is estimated to be just as desirable as a chance of winning the bigger prize with
probability p, we may assign a utility 1/p to the bigger prize (and this, of course, is larger
than 1). Similarly, the original "ticket" scale may have to be extended to assign negative
utilities to certain undesirable situations. Considering such a situation "in context", as an
outcome of a prospect whose other outcomes are quite positive, allows the semi-direct
use of the "ticket" scale to evaluate its negative utility.
In this paper we apply capability index as a utility function. A common Process
Capability measure, Cp (often called a Process Capability Index), indicates how well the
process distribution fits within its specification limits, and is simply the ratio of the
specification width to the variation width. Process capability compares the output of an
in-control process to the specification limits by using capability indices. The comparison
is made by forming the ratio of the spread between the process specifications (the
specification "width") to the spread of the process values, as measured by 6 process
standard deviation units (the process "width").

σ 6
LSL USL
C
p

= ,

where USL is upper specification limit and LSL is lower specification limit. The σ is
variation width of the process. USL and LSL in this paper would be the standards which
The 3
rd
International Conference on Virtual Learning, ICVL 2008

139
the educational standard institute identifies based on the VLEs facilities, quality of
education, user satisfaction and etc.
We are often required to compare the output of a stable process with the process
specifications and make a statement about how well the process meets specification. To do
this we compare the natural variability of a stable process with the process specification
limits. A capable process is one where almost all the measurements fall inside the
specification limits. This can be represented pictorially by the plot below (Figure 2):



Figure 2. A Capable Process


4. Integer Programming Approach

Here we introduce the mathematical model for deciding about the optimal
curriculum for a student. The proposed model is based on 0/1 integer programming. The
indexes, notations, and equations are described as follows:
University of Bucharest and Ovidius University of Constanta


140

Indexes:
i Variable for the number of departments
j Variable for the number of courses
k Variable for the number of the professors
Notations:
s
ij

The number of teachers who offer the course j in department i
D
i
Whole courses that can be chosen in one department.
l The total number of units that the student should pass when they are graduated
m The number of departments
n The number of courses
Decision variable:
X
ijk
1 If Student choose department i and, take the course j by professor k; 0 if not

The mathematical model:


∑∑∑
= = =
× =
n
i
m
j
S
K
ijk p
ij
X C Z
1 1 1
max (1)
S.T.
For j = 1,2,…,m
∑∑
= =

n
i
S
k
ijk
ij
X
1 1
1 (2)


∑∑∑
= = =
=
n
i
m
j
S
k
ijk
ij
l X
1 1 1
(3)

For i=1,2,….,n
∑∑
= =
× =
m
j
S
k
i ijk
ij
D l X
1 1
(4)

l D l
n
i
i
= ×

=1
(5)


ijk p ijk
X C X × ≥ + 33 . 1

(6)

X
ijk
, D
i
= 0,1 (7)

Equation (1), which is the objective function, wants to reach the optimal curriculum
based on the utility function. Equation (2) guaranties that each course is not selected or
just one time is selected. Equation (3) confines the number of course selection to l.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

141
Equation (4) and (5) guaranty that all selected courses by a student should be in one
educational center. Equation (6) indicates that if a course offered by an educational center
is not of suffice utility, then it automatically won’t be selected by a student. Equation (7)
presents the sign and the kind of the variables.
In this way after some computations the optima curriculum is achieved. The curriculum
consists of the department, the course, and teacher. The advantages of the proposed
model is excluding personal preferences and replace it with capability index of education
which would be same for all students, providing a mathematical model structure for the
problem which reduce the uncertainty in decision modeling, and provision of effective
educational elements for curriculum selection regarding to student’s profile.


5. Conclusions

Virtual Learning system is an internet based service like the application system or
the internet based virtual course study service. This system is able to be interpreted in
various ways of learning. One of the aspects of virtual learning environments is curriculum
selection. Most of the time students would choose their curriculum based on their preferences
but the preference won’t all the time provide correct selection considering different
emotions of human. Therefore in this paper we have applied a standard for selecting
courses, teachers, and departments that is called utility function for decision maker.
Process capability is a standard that is computed to determine the quality of an element.
Integrating that capability index with 0/1 integer programming we achieved the optimal
curriculum for a student. As future study we will develop the model for multi student
associated with stochastic time period each student may enter into the selecting process.


6. Acknowledgments

This research has been supported by Mazandaran Telecommunication Research center.



REFERENCES


BIEGLER, L. T., GROSSMANN, I. E., 2004. Retrospective on optimization. Computers and
Chemical Engineering 28, 1169-1192.
DJERDJOUR, M., 1997, An enumerative algorithm framework for a class of nonlinear integer
programming problems. European Journal of perational Research 101, 104-121.
University of Bucharest and Ovidius University of Constanta


142
PICCOLI, G., AHMAD, R., & IVES, B. (2001), Web-Based Virtual Learning Environments: A
Research Framework and a Reliminary Assessment of Effectiveness in Basic IT Skills
Trening, MIS Quarterly, 25(4), 401-426.
SALKIN, H. M., MATHUR, K., 1989, Foundations of Integer Programming. North-Holland, Amsterdam.
SIMONS, R. 1996, How OR Improved Smelter Performance. MP in Action. The Newsletter of
Mathematical Programming in Industry and commerce, United Kingdom, February.
SONG, Y. S. (2000), Cultivation course to men of talent and e-Learning strategy of digital era.
Journal of Training and Development, 2000(7), 148.
TAHA, H. A., 2003, Operations Research. Seventh ed. Macmillan, New York.
WANG, Y. S. (2003), Assessment of learner satisfaction with asynchronous electronic learning
systems. Information & Management, 41(1), 75.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

143
Selection of Optimum Maintenance Strategies in a Virtual
Learning Environment based on Analytic Hierarchy Process

Hamed Fazlollahtabar
1
*, Narges Yousefpoor
2

1
Young Researchers Club, Islamic Azad University Babol Branch, Iran
2
Department of Industrial Engineering Mazandaran
University of Science and Technology, Babol, Iran
*E-mail: hamed@http.ac.ir


Abstract
This paper aims to evaluate different maintenance strategies (such as corrective
maintenance, time-based preventive maintenance, condition-based maintenance, and
predictive maintenance) for different equipment used in a virtual learning
environment. An optimal maintenance strategy mix is necessary for increasing
availability and reliability levels of learning facilities without a great increasing of
investment. The selection of maintenance strategies is a typical multiple criteria
decision-making (MCDM) problem. Therefore Analytic Hierarchy Process (AHP) is
applied to select the appropriate maintenance strategy.

Keywords: Maintenance strategies; virtual learning environment; Multiple criteria
decision-making; analytic hierarchy process.


1. Introduction

The capability and flexibility of the virtual learning system (VLS) have been
demonstrated in both training and education, resulting in its adoption by the academia as
well as the industry. Since the commercial application package (or commercial off-the-shelf)
strategy of system development is so widespread (Whitten, Bentley, & Dittman, 2004),
the proliferation of VLS applications has created confusion for the potential adopters when
they have to make a decision regarding the selection from candidate products or solutions.
Conventional approaches for evaluating an information system (IS) have leaned
towards the standpoints of technical personnel. In contrast, the VLS places particular stress on
certain areas, such as, the content and the ways in which it is presented, demonstrating
that it is a highly user-oriented system. Since users are widely recognized as the key
stakeholders in any IS or IS service (Jiang, Klein, Roan, & Lin, 2001), their attitudes
toward the system are pivotal and should be valued. This is evidenced by the fact that
user satisfaction is often seen as a key antecedent to predict the success of a particular IS
(DeLone & McLean, 2003), or to anticipate a user’s behavior of reuse (Lin & Wang, 2006;
Lin, Wu, & Tsai, 2005).
Unfortunately, unlike other aspects of administering a learning institute which have
received tremendous interest from researchers and practitioners, maintenance received
little attention in the past. This is one of the reasons that results in low maintenance
University of Bucharest and Ovidius University of Constanta


144
efficiency in learning institute at present. As indicated by Mobley (2002), one third of all
maintenance costs is wasted as the result of unnecessary or improper maintenance
activities. Today, research in this area is on the rise. Moreover, the role of maintenance is
changing from a ‘‘necessary evil’’ to a ‘‘profit contributor’’ and towards a ‘‘partner’’ of
institutes to achieve world-class competitiveness (Waeyenbergh and Pintelon, 2002).
Therefore, research on maintenance represents an opportunity for making significant
contribution by academics.
In the literature, maintenance can be classified into two main types: corrective and
preventive (Li et al., 2006; Waeyenbergh and Pintelon, 2004). Corrective maintenance is
the maintenance that occurs after systems failure, and it means all actions resulting from
failure; preventive maintenance is the maintenance that is performed before systems
failure in order to retain equipment in specified condition by providing systematic
inspections, detection, and prevention of incipient failure (Wang, 2002). Based on the
development of preventive maintenance techniques, three divisions of preventive
maintenance are considered in this paper, i.e. time-based preventive maintenance,
condition-based maintenance, and predictive maintenance.
Therefore, the MCDM theory should be used for the maintenance strategy
selection. Several MCDM methods have been developed, such as the weighted sum
model (WSM), the weighted-product model (WPM), the TOPSIS method, and the AHP
(Triantaphyllou and Lin, 1996). The AHP is one of the most popular MCDM methods. It
has the following advantages (Triantaphyllou et al., 1997; Bevilacqua and Braglia, 2000):
(1) It is the only known MCDM model that can measure the consistency in the
decision makers’ judgments;
(2) The AHP can help the decision makers to organize the critical aspects of a
problem into a hierarchical structure similar to a family tree, making the decision process
easy to handle;
(3) Pair-wise comparisons in the AHP are often preferred by the decision makers,
allowing them to derive weights of criteria and scores of alternatives from comparison
matrices rather than quantify weights/ scores directly.
In this paper, first we describe different maintenance strategies. Then the criteria
and sub-criteria are defined and finally, an AHP-based decision model is proposed.


2. Alternative Maintenance Strategies

Four alternative maintenance strategies considered in this paper are introduced as
following:
(1) Corrective maintenance: This alternative maintenance strategy is also named as
fire-fighting maintenance, failure based maintenance or breakdown maintenance.
(Swanson, 2001). Corrective maintenance is the original maintenance strategy appeared
in industry (Waeyenbergh and Pintelon, 2002; Mechefske and Wang, 2003). It is considered as
a feasible strategy in the cases where profit margins are large (Sharma et al., 2005).
(2) Time-based preventive maintenance: According to reliability characteristics of
equipment, maintenance is planned and performed periodically to reduce frequent and
The 3
rd
International Conference on Virtual Learning, ICVL 2008

145
sudden failure. For performing time-based preventive maintenance, a decision support
system is needed, and it is often difficult to define the most effective maintenance
intervals because of lacking sufficient historical data. In many cases when time-based
maintenance strategies are used, most equipments are maintained with a significant
amount of useful life remaining (Mechefske and Wang, 2003).
(3) Condition-based maintenance: Maintenance decision is made depending on the
measured data from a set of sensors system when using the condition-based maintenance
strategy. But limitations and deficiency in data coverage and quality reduce the
effectiveness and accuracy of the condition-based maintenance strategy (Al-Najjar and
Alsyouf, 2003).
(4) Predictive maintenance: In the literature, predictive maintenance often refers to
the same maintenance strategy with condition-based maintenance (Sharma et al., 2005;
Mobley, 2002). In this paper, considering the recent development of fault prognosis
techniques (Bengtsson, 2004; Byington et al., 2002), predictive maintenance is used to
represent the maintenance strategy that is able to forecast the temporary trend of
performance degradation and predict faults of equipments by analyzing the monitored
parameters data.
Recently, the intelligent maintenance system was described by Djurdjanovic et al.
(2003), focusing on fault prognostic techniques and aiming to achieve near-zero-
downtime performance of equipment. However, generally speaking, the amount of
equipment failure can be reduced if the preventive maintenance strategies are correctly
selected, especially the condition-based/predictive maintenance.


3. Comparing Criteria

When different maintenance strategies are evaluated for different equipments, the
learning institutes must set maintenance goals taken as comparing criteria first. Different
learning institutes may have different maintenance goals. But in most cases, these goals
can be divided into four aspects analyzed as follows:
(1) Safety: Safety levels required are often high in many learning institutes. The
relevant factors describing the Safety are:
(a) Personnel: The failure of many types of equipment can lead to serious
damage of personnel on site.
(b) Facilities: For example, the sudden breakdown of a learning server in a
virtual learning system can result in serious damage of the corresponding
learners who are being serviced by that server.
(c) Environment: The failure of equipment with poisonous liquid or gas can
damage the environment.
(2) Cost: Different maintenance strategies have different expenditure of hardware,
software, and personnel training.
(a) Hardware: For condition-based maintenance and predictive maintenance, a
number of sensors and some computers are indispensable.
University of Bucharest and Ovidius University of Constanta


146
(b) Software: Software is needed for analyzing measured parameters data when
using condition- based maintenance and predictive maintenance strategies.
(c) Personnel training: Only after sufficient training can maintenance staff make
full use of the related tools and techniques, and reach the maintenance goals.
(3) Added-value: A good maintenance program can induce added-value, including
low inventories of spare parts, small production loss, and quick fault identification.
(a) Spare parts inventories: Generally, corrective maintenance need more spare
parts than other maintenance strategies. Spare parts for some equipments
are really expensive.
(b) Production loss: The failure of more important equipments in the learning
process often leads to higher production loss cost. Selecting a suitable
maintenance strategy for such equipments may reduce production loss.
(c) Fault identification: Fault diagnostic and prognostic techniques involved in
the condition- based and predictive maintenance strategies aim to quickly
tell maintenance engineers where and why fault occurs. As a result, the
maintenance time can be reduced, and the availability of the learning
system may be improved.
(4) Feasibility: The feasibility of maintenance strategies is divided into acceptance
by users and technique reliability.
(a) Acceptance by users: Managers and maintenance staff prefer the
maintenance strategies that are easy to implement and understand.
(b) Technique reliability: Still under development, condition-based maintenance
and predictive maintenance may be inapplicable for some complicated
learning facilities.


4. Analytic Hierarchy Process

The AHP was developed first by Satty. It is a popular tool for MCDM by
structuring a complicated decision problem hierarchically at several different levels. The
algorithm for applying AHP method in the proposed problem of this study is as follows:


Notational Definitions:

n: Number of criteria; m: Number of strategies; p: Index for strategies, p = 1,2,…,m.
j: Index for sub-criteria, j = 1,2,…,11; q: Index for criteria, q = 1,2,…,4.
p, j
W′ : The weight of pth strategy with respect to jth sub-criterion.
j ,q
W : The weight of jth sub-criteria with respect to qth criterion.
p,q
R : The weight of pth strategy with respect to qth criterion.
q
w : The weight of qth criterion.

The following steps are taken for the proposed problem of this study:
The 3
rd
International Conference on Virtual Learning, ICVL 2008

147
Step 1: Define the decision problem and goal.
Step 2: Structure the hierarchy from the top through the intermediate to the lowest
level. The hierarchy structure for the proposed problem is shown in Figure 1.

Goal
Safety
Added-value
Cost
Feasibility
Facilities
Environment
Personnel
Production Loss
Spare Part Inventories
Fault Identification
Hardware
Software
Personnel Training
Acceptance by Labors
Technique Reliability
Predictive Maintenance
Condition-based
Maintenance
Time-based Predictive
Maintenance
Corrective Maintenance


Figure 1. Hierarchy Structure for the Proposed Problem

Step 3: Construct the strategy-criteria matrix using steps 3-1 to 3-5 by the AHP method.
Step 3-1: Matrices of pair-wise comparisons are constructed for each of the
lower levels with one matrix for each element in the level immediately
above by using a relative scale measurement. The decision maker has
the option of expressing his or her intensity of preference on a nine-
point scale. If two criteria are of equal importance, a value of 1 is given
in the comparison, while a 9 indicates an absolute importance of one
criterion over the other.
Step 3-2: Computation of eigenvalue by the relative weights the criteria and the
sum is taken over all weighted eigenvector entries corresponding to
those in the next lower level of the hierarchy.
Pair-wise comparison data can be analyzed using the eigenvalue
technique. Using these pair-wise comparisons, the parameters can be
estimated. The right eigenvector of the largest eigenvalue of matrix A
constitutes the estimation of relative importance of attributes.
Step 3-3: Construct the consistency and consequence weights analysis, as
follows:
University of Bucharest and Ovidius University of Constanta


148

( )
1 1
2
2 2
1
1 2
1
1
1
n
n
ij
n n
w w
w w
w w
w w
A a
w w
w w




= =





K
K
M M O


Note that if matrix A is consistent (that is,
ij ik kj
a a a = for all , , 1, 2, ..., i j k n = ),
then A contains no errors (the weights are already known) and we have

, , 1, 2, ...,
i
ij
j
w
a i j n
w
= =
.

If the pair-wise comparisons do not include any inconsistencies, than
max
λ = n. The
more consistent the maximum comparisons are, the closer the value of computed
max
λ is
to n. A consistency index (CI), which measures the inconsistencies of pair-wise
comparisons, is set to be:

( )
( )
max
1
n
CI
n
λ −
=

, and a consistency ratio (CR) is set to be: 100 ,
CI
CR
RI
| |
=
|
\ ¹


where n is the number of columns in A, CI is the consistency index, and RI is the random
index, being the average of the CI obtained from a large number of randomly generated matrices.
Note that RI depends on the order of the matrix, and CR value of 10% or less is
considered acceptable.
Steps 3-1 to 3-3 are performed for all levels in the hierarchy.
Step 3-4: Configure strategy-sub criteria and the sub criteria-criteria matrix as follow:

Table 1
The strategy-sub criteria matrix

SC
1
SC
2
… SC
11

Strategy 1 W'
1,1
W'
1,2
… W'
1,11

Strategy 2 W'
2,1
W'
22
… W'
2,11

Strategy 3 W'
3,1
W'
3,2
… W'
3,11

Strategy 4 W'
4,1
W'
4,2
… W'
4,11

The 3
rd
International Conference on Virtual Learning, ICVL 2008

149




Table 2
The sub criteria-criteria matrix

C1 C2 C3 C4
SC
1
W
1,1
W
1,2
W
1,3
W
1,4

SC
2
W
2,1
W
22
W
2,3
W
2,4


SC
11
W
11,1
W
11,2
W
11,3
, W
11,4


Step 3-5: The strategy-criteria matrix is formed as follows:


Table 3
The strategy-criteria matrix

C
1
C
2
C
3
C
4

Strategy 1 R
11
R
12
R
13
R
14

Strategy 2 R
21
R
22
R
23
R
24

Strategy 3 R
31
R
32
R
33
R
34

Strategy 4 R
41
R
42
R
43
R
44



where 1,..., , 1,..., 4
pq pj jq
j
R W W p m q ′ = × ∀ = =

, and j is the number of sub criteria.
Step 4: As result we can configure the pair-wise comparison for criteria-criteria
matrix in Table 4.




University of Bucharest and Ovidius University of Constanta


150
Table 4
The criteria-criteria pair-wise comparison matrix

C
1
C
2
C
3
C
4
w
q
Criteria 1 1 a
12
a
13
a
14
w
1
Criteria 2 1/a
12
1 a
23
a
24
w
2
Criteria 3 1/ a
13
1/a
23
1 a
34
w
3

Criteria 4 1/a
14
1/a
24
1/a
34
1 w
4


The w
q
s are gained by normalization process. Hence the w
q
s are the weights for criteria.
Step 5: Here we can calculate the overall weights for the strategies using Tables 3
and 4, as follows:
4 44 3 43 2 42 1 41
4 34 3 33 2 32 1 31
4 24 3 23 2 22 1 21
4 14 3 13 2 12 1 11
4 strategy for weight Total
3 strategy for weight Total
2 strategy for weight Total
1 strategy for weight Total
w R w R w R w R
w R w R w R w R
w R w R w R w R
w R w R w R w R
× + × + × + × =
× + × + × + × =
× + × + × + × =
× + × + × + × =

Here we achieve the weights for the alternatives (strategies), hence the ranking is
possible. The strategy which receives the highest weight is number one in ranking and
therefore, is the strategy that will be selected.


5. Conclusion

In this paper, the selection of maintenance strategies in virtual learning institutes is
studied. An optimal maintenance strategy mix can improve availability and reliability
levels of plants equipment, and reduce unnecessary investment in maintenance. The
evaluation of maintenance strategies for each piece of equipment is a multiple criteria
decision-making (MCDM) problem. Considering the nature of decision making, the AHP
method is used for the evaluation of different maintenance strategies. The AHP models
the problem in hierarchy and evaluate the criteria, sub-criteria and alternatives
thoroughly. The final result of AHP is an overall ranking of alternatives. The proposed
approach of this study is useful for other similar MCDM problems.


6. Acknowledgments

This research has been supported by Mazandaran Telecommunication Research center.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

151
REFERENCES


AL-NAJJAR, B., ALSYOUF, I., 2003, Selecting the most efficient maintenance approach using fuzzy multiple
criteria decision making. International Journal of Production Economics 84, 85-100.
BENGTSSON, M., 2004, Condition based maintenance system technology-where is development heading?
Proceedings of the 17th European Maintenance Congress, Barcelona, Spain, May 11-13.
BYINGTON, C., ROEMER, M. J., GALIE, T., (2002), Prognostic enhancements to diagnostic systems for
improved condition- based maintenance. Proceedings of IEEE Aerospace Conference, vol. 6, Big Sky,
USA, March 9-16, pp. 2815-2824.
DELONE, W. H., & MCLEAN, E. R. (2003), The DeLone and McLean model of information systems
success: a ten-year update. Journal of Management Information Systems, 19(4), 9-30.
DJURDJANOVIC, D., LEE, J., NI, J., (2003), Watchdog agentan infotronics based prognostics approach for
product performance assessment and prediction. International Journal of Advanced Engineering
Informatics 17, 109-125.
JIANG, J. J., KLEIN, G., ROAN, J., & LIN, J. T. M. (2001), IS service performance: self-perceptions and user
perceptions. Information & Management, 38(8), 499-506.
LI, J. R., KHOO, L. P., TOR, S. B., (2006), Generation of possible multiple components disassembly sequence
for maintenance using a disassembly constraint graph. International Journal of Production Economics
102, 51-65.
LIN, C. S., WU, S., & TSAI, R. J. (2005), Integrating perceived playfulness into expectation-confirmation
model for web portal context. Information & Management, 42(5), 683-693.
LIN, H.-H., & WANG, Y.-S. (2006), An examination of the determinants of customer loyalty in
mobile commerce contexts. Information & Management, 43(3), 271-282.
MECHEFSKE, C. K., WANG, Z., 2003, Using fuzzy linguistics to select optimum maintenance and
condition monitoring strategies. Mechanical Systems and Signal Processing 17 (2), 305-316.
MOBLEY, R. K., 2002, An Introduction to Predictive Maintenance, Second ed. Elsevier science,
New York.
SHARMA, R. K., KUMAR, D., KUMAR, P., 2005, FLM to select suitable maintenance strategy
in process industries using MISO model. Journal of Quality in Maintenance Engineering 11
(4), 359-374.
SMITH, C. U., & WILLIAMS, L. G. (1999), A performance model interchange format. Journal of
Systems and Software, 49(1), 63-80.
SWANSON, L., 2001, Linking maintenance strategies to performance. International Journal of
Production Economics 70, 237-244.
TRIANTAPHYLLOU, E., KOVALERCHUK, B., MANN, L., KNAPP, G. M., 1997, Determining the
most important criteria in maintenance decision making. Journal of Quality in Maintenance
Engineering 3 (1), 16-28.
TRIANTAPHYLLOU, E., LIN, C. T., 1996, Development and evaluation of five fuzzy
multiattribute decision-making methods. International Journal of Approximate Reasoning
14, 281-310.
WAEYENBERGH, G., PINTELON, L., 2002, A framework for maintenance concept development,
International Journal of Production Economics 77, 299-313.
University of Bucharest and Ovidius University of Constanta


152
WAEYENBERGH, G., PINTELON, L., 2004, Maintenance concept development: A case study.
International Journal of Production Economics 89, 395-405.
WANG, H., 2002, A survey of maintenance policies of deteriorating systems. European Journal of
Operational Research 139, 469-489.
WHITTEN, J. L., BENTLEY, L. D., & DITTMAN, K. C. (2004), System analysis & design methods,
New York: McGraw-Hill.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

153
Applying QFD Approach to Design an Online Course
in a Virtual Learning Environment

Hamed Fazlollahtabar

Young Researchers Club, Islamic Azad University Babol Branch, Iran
e-mail: hamed@ustmb.ac.ir


Abstract
Migration from the traditional to web-based learning paradigm is usually
accompanied by remodeling of many learning core activities particularly those
associated with user-centered services. In this capacity of the web-based learning
paradigm, many educational centers have established networked environments
within which many virtual-user communities are forming and growing.
Understanding the virtual user’s needs in these communities has become the first
priority of networked learning systems for designing, running and managing
effective virtual learning services to meet the increasing expectations of the invisible
users. To achieve this, the virtual learning system strives to improve their quality of
service by applying a wide range of such quality management approaches as quality
function deployment (QFD). QFD initially stresses on driving continuous
improvement of the user-oriented services towards end-user satisfaction. The paper
attempts to incorporate the QFD to be integrated strategically in designing and
managing e-learning provision within networked learning environment.

Keywords: virtual learning environment, Quality function development,
user-oriented services.


1. Introduction

Perhaps no other factor in the history of the information and communication
technology (ICT) has changed the face of information use and delivery as significantly as
the swift emergence of the Internet and related web-based applications in academic
settings. The evolving genre of the web has engendered new paradigm of e-research
community whose hallmark is scholarly use of the web within ICT-rich learning
environments. The significant product of the ICT revolution is the networked education.
The networked education has remarkably extended the breadth and scale of scholarly
evidence to support innovative learning and research activities.
Migration from the traditional to the web-based education paradigm is usually
accompanied by remodeling of many core activities of the networked learning
particularly those associated with the user-centered services. Networked learning
functions and academic information requirements are inextricably linked. This statement
can be translated into: i – quality of information services; ii – efficiency of delivery
system; and iii – satisfaction of information consumers. These components motivated
University of Bucharest and Ovidius University of Constanta


154
educational centers in incorporating a wide range of quality management approaches (e.g.
quality function deployment (QFD)) as an effective means of incorporating quality
improvement in their user-centered information services. Networked learning services
can be illustrated as open interrelated systems with input-output interoperability where
the education administration should maintain user-oriented collection development and
information commons as an input and end-user satisfaction as output (Hsieh; et al., 2000).
In recent years, the appeal of web-based instruction has grown dramatically. The
idea of having 24-hour access to curriculum from any remote location has students
searching for institutions that offer online courses in their discipline. Fields and
Huffstutter (2000) state, “To students, the key benefit of such virtual offerings is
flexibility and time. Students can log in at their leisure, email their papers and post notes
to classmates whenever they want”. Now that the demand is rising and online courses are
becoming more commonplace, institutions are responding by offering more classes
online. In fact, Charp (2002) indicated that the number of online courses and colleges will
increase drastically in near future. However, developing and delivering an effective
online course requires specialized training, proper planning, and significant amounts of
time. In an analysis presented by Palloff & Pratt (1999), it showed roughly seven hours
per week for a face-to-face class and approximately 18 hours per week for an online
class. The use of Quality Function Deployment (QFD) in the initial design process can
prove beneficial and save valuable time in the course design process. The transformation
of a face-to-face course to a web-based course is not a simple process. Saving PowerPoint
slides or class notes as HTML and posting them on the web does not qualify as an
instructionally sound web course. Meaning, faculty should adopt an instructional design
paradigm that moves away from instructor-controlled (systematic) learning and toward
learner-centered (dynamic) instruction. Because of the availability and focus of new
technologies faculty should be facilitating online experiences that help learners become
skilled at finding and accessing information, evaluating it critically, and using it to solve
problems (Gillespie, 1998).
Quality Function Deployment (QFD), Akao “the voice of the customer” is a
problem prevention tool. This model is a systematic method for structured product
planning and development that enables developers to clearly identify customers’
(students’) wants and needs, and then evaluate each proposed component or service
capability systematically in terms of its impact on meeting the expressed desires of the
customer (Stein, 2002). QFD was conceived by Yoji Akao during the late 60’s in Japan.
However, it was not until 1972 that QFD was publicly recognized when applied at the
Mitsubishi shipyards in Japan. QFD was first introduced by two interrelated objectives
(Akao, 1972, 1997), these were:
• To convert the core desire, demand and need of the end-users for interesting products
into substitute quality characteristics (SQC) at different stages of design and testing.
• To assure that SQC is properly deployed throughout the processes of
manufacturing, production, and delivery of new products or services.
If the producer succeeded in bringing the two objectives together, its product would
meet the satisfaction of the end-users (Han et al., 2001; ReVelle et al., 1997). Moreover,
we can view QFD as a fundamental trade-off between the end-users and the producers.
The QFD has been experienced a vast range of development and modifications to yield a
The 3
rd
International Conference on Virtual Learning, ICVL 2008

155
rigorous analytic tool to understand end-user behavior for developing comprehensive
product and service specifications through creating end-user strategies and developing a
mechanism for enabling such strategies (Killen et al., 2005). With it roots originally
planted in the industrial sectors, QFD has now found acceptance in departmental research
in education. These applications range from textbook selection to redesign of
departmental business operations. Regardless of the application, the QFD process consists
of four primary areas of focus (see Figure 1). The details of each component go beyond
the scope of this paper and should be investigated as a separate issue for faculty
unfamiliar with QFD. However, the general QFD application focus areas and their
operational definitions for this paper are as follows:
1. Product planning – content and audience analysis
2. Part deployment – development of course objectives
3. Process planning – course activities and instructional methods
4. Production planning – delivery techniques

Product Planning
Process Planning
Parts Development
Production Planning
How much
How much
How much
How much
S
t
u
d
e
n
t

E
x
p
e
c
t
a
t
i
o
n
s
Content/Audience
Analysis
C
o
n
t
e
n
t
/
A
u
d
i
e
n
c
e

A
n
a
l
y
s
i
s
Course Objectives
C
o
u
r
s
e

O
b
j
e
c
t
i
v
e
s
Course Activities/
Instructional Techniques
C
o
u
r
s
e
A
c
t
iv
it
i
e
s
/
I
n
s
t
r
u
c
ti
o
n
a
l
T
e
c
h
n
iq
u
e
s
Delivery Techniques


Figure 1. Quality Function Deployment Stages


2. Product planning stage

The initial phase of the QFD is the product planning stage. During this stage the
purpose is to acquire students’ input to define the characteristics of a quality online
course from their perspective. Obtaining student information can be accomplished
through personal interviews, focus groups, telephone calls, surveys (online and paper), or
whatever methods are available. The primary goal is to elicit feedback from those who
have taken an online course and input from those who are interested in taking an online
course. In addition, faculty must analyze course goals and content to determine course
objectives. Product planning is the most critical and difficult step of the process.
Maintaining objectivity and capturing the essence of the students’ needs and expectations
is vital to ensuring a successful e-learning experience while analyzing course content is
the first step in ensuring that the course is instructionally sound.
University of Bucharest and Ovidius University of Constanta


156
In trying to achieve the ideal e-learning experience faculty should begin by asking
the following types of questions:
1. Who are the students? Is this course a requirement for them or an elective? Are
they academically mature?
2. What are their needs?
3. Where does this course fit into the curriculum?
4. What content should be taught in this course?
After the data is collected and categorized through the use of affinity diagrams the
process moves toward the creation of the initial House of Quality (HOQ) (see Figure 2).
This portion of the process will allow the course designer to transform the information
into quantitative data that is useful in the analysis and prioritization of course elements.
The course elements that gain top priority are then shifted to the next House of Quality to
begin the part deployment stage.

Technical Correlations
Technical Response
C
u
s
t
o
m
e
r

N
e
e
d
s

a
n
d

B
e
n
e
f
i
t
s
P
l
a
n
n
i
n
g

M
a
t
r
i
x
Technical Matrix (Technical Response Priorities)
Relationships
(Impact of Technical Response on Customer Needs and Benefits)


Figure 2. House of Quality


3. Part deployment stage

During the part deployment phase faculty are required to establish course
objectives to ensure the course meets curricular requirements and is instructionally sound.
In doing so, faculty should continue to analyze course content and student needs. At this
point in the process, faculty must be careful not to compromise the instructional integrity
of the course. The writing of objectives and test items is not to be taken lightly.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

157
Objectives are the infrastructure upon which instructional experiences are built. However,
proper writing of objectives goes far beyond the scope of this paper and therefore could
not be addressed thoroughly in this format. Faculty members are encouraged to pursue
additional guidance on the writing of objectives in order to ensure it is done properly. The
following types of questions might be used to initiate the process.
1. What are the course objectives?
2. What instructional activities will facilitate the meeting of course objectives?
3. What types of test items, or other means of evaluation, will be used to
determine whether, or not, students master the content and meet course objectives?


4. Process planning stage

Process planning is used to focus on the technical operations of the online course.
Faculty has the opportunity to prescribe the flow and delivery styles of information and
activities. For example, an instructor presenting lecture material on hiring front-line
supervisors may choose to deliver a PowerPoint lecture followed by streaming video and
an online discussion of the topic. Another instructor delivering the same lecture may
choose to remove the PowerPoint slides and inject an online quiz relating to an
appropriate chapter in the textbook. The ways in which faculty delivery their course
materials and select course activities is unique to their teaching style. However, through
the use of QFD we can structure the learning processes to best suite our students.
As the course development process begins, faculty must consider the style of web-based
instruction and the learning styles of the students. Due to the diversity of student learning
styles and academic maturity it is beneficial for instructors to utilize several different
instructional approaches and activities. The acknowledgement of this incongruence should
serve as a design consideration to avoid producing dysfunctional web-based courses.
Although it is humanly impossible to meets the needs of all students in face-to-face
environments, technology has made it possible to offer several varieties of learning
activities to engage students in a web-based environment.
Developing a variety of activities to address students’ needs can prove to be a
formidable task. There are multiple modes of delivery (text-based, streaming video/audio,
graphic interfaces, etc.) available through the Internet, however all students are not
equipped with the same hardware and software. Therefore, activities must be generated to
work with a set of minimal standards (hardware and software) that are available to all
students. This can be achieved by using programs that offer shareware to students.
Examples of such software are Macromedia Flash, Adobe Acrobat, Real Media Player,
and PowerPoint viewer. Providing students multiple options for learning course materials
has proven to be effective. Additionally, the availability of options has the potential to
facilitate an opportunity to help students learn or acquire information in alternate
methods. In general, the choice of media for any learning activity will of course be
heavily influenced by the range of issues related to how and when the activities will be
undertaken by the students (Ryan et al., 2000).
University of Bucharest and Ovidius University of Constanta


158
In order to ensure that the quality of a course is enhanced by the activities, faculty
must address the following:
1. Which course objective is addressed by each activity?
2. What, exactly will students gain from engaging in the prescribed activities?
3. Is the value of the activities obvious to students?
As course content and activities will vary, instructors are encouraged to continue
the questioning process as needed for their online course(s). Overall, online activities
must support course objectives in a manner that is meaningful to students and
instructionally sound. An additional feature that was critical to Southeast choosing to
create and use OIS was the clean architecture of the internal and external communication
functions. In a web-based learning environment communication between the instructor, the
software, and students is vital. Being able to maintain a high-level of secure communication
throughout the e-learning experience is essential for academic success. By moving all
communications modes (email, file sharing, electronic posts, etc.) inside the software,
faculty and students are provided a heightened sense of privacy and security.


5. Production planning stage

The last stage of the QFD process is the production planning. The goal of the
production planning stage is to outline and structure the material, activities, and
technologies necessary to deliver the course online. The variety of online course delivery
systems currently available allows institutions to choose from several options. As the
market continues to grow with the advancements in technology, the selection of a
software package can be a daunting task. In a web-based learning environment
communication between the instructor, the software, and students is vital. Being able to
maintain a high-level of secure communication throughout the e-learning experience is
essential for academic success. By moving all communications modes (email, file
sharing, electronic posts, etc.) inside the software, faculty and students are provided a
heightened sense of privacy and security.
Technology selection and support has a significant impact on teaching online, both
for faculty and students. Therefore, consideration must be given to the relationship
between instructional needs and the available technological capabilities before venturing
into the e-learning arena. At a minimum, faculty should remember that online courses
require reliable, scalable, and flexible information technology capabilities for communicating,
collaborating, and information sharing (Rai, 2000).


6. The other final stage (evaluation)

Although the production planning stage is referred to as the last stage this is not
actually the case where course design and development is concerned. Upon completion of
the production planning stage faculty should have a working prototype of the course. As
with all prototypes, a thorough review of the online course should be performed before it
The 3
rd
International Conference on Virtual Learning, ICVL 2008

159
is deployed. A good method of evaluation would be to use a small group of students to
pilot the course as an alternate delivery to a current face-to-face course. The feedback
from these students will be necessary to see if the course design is effective. Once a beta-
test of the course has been conducted the course will be either revised or packaged for use.


7. Conclusions

Well-developed online courses that facilitate above-average learning experiences
are not commonplace. They are overshadowed by the multitude of self-directed online
correspondence courses. As faculty set out to design an online course they should give
thought to the entire process, from development to deployment. Otherwise, they can
create frustration for themselves and their students. Quality Function Deployment (QFD)
can be useful in the course development process. It is a simple, yet powerful, means of
discovering key characteristics of a successful online course. Being proactive and using
QFD principles properly will help faculty to identify instructional design and technical
concerns early in the design process. Utilizing QFD early in the planning stages of an
online course will minimize frustration and maximize the learning process. Both of which
can lead to a strong e-learning experience for students and faculty.
The use of Quality Function Deployment and contemporary instructional design
models can assist in the successful design of a web-based course. Understanding the
student’s expectations and desires is critical.


8. Acknowledgments

This research has been supported by Mazandaran Telecommunication Research Center.



REFERENCES


AKAO, Y. (1972), New product development and quality assurance: Quality deployment system
(translated from Japanese). Standardization and Quality Control, 25(4), 7-14.
AKAO, Y. (1997), QFD: past, present, and future. In Proceedings of the 3rd Annual International Symposium
on QFD (QFD’97) [Online]. Retrieved 9 December 2005 from http://www.qfdi.org/QFD_History.pdf
AKAO, Y., & MAZUR, G. H. (2003), The leading edge in QFD: past, present, and future.
International Journal of Quality & Reliability Management, 20(1), 20-35.
CHARP, S. (2002, March), Online learning. THE Journal, 29(8), 8-9.
FIELDS, R. & HUFFSTUTTER, P. J. (2000, March 3), A virtual revolution in teaching. The Los
Angeles Times, p. A.1.
GILLESPIE, F. (1998, Winter), Instructional Design for the New Technologies, New Directions
for Teaching & Learning, 76, 39-52.
HAN, L., & GOULDING, A. (2003), Information and reference services in the digital library.
Information Services & Uses, 23(4), 251-262.
University of Bucharest and Ovidius University of Constanta


160
HAN, S. B., CHEN, S. K., EBRAHIMPOUR, M., & SODHI, M. S. (2001), A conceptual QFD
planning model. International Journal of Quality & Reliability Management, 18(8), 796-718.
HSIEH, P-N., CHANG, P-L., & LU, K-H. (2000), Quality management approaches in library and
information services. Libri, 50(3), 191-201.
KILLEN, C. P., WALKER, M., & HUNT, R. A. (2005), Strategic planning using QFD.
International Journal of Quality & Reliability Management, 22(1), 17-29.
PALLOFF, R. M. & PRATT, K. (1999), Building learning communities in cyberspace, San
Francisco: Jossey-Bass.
RAI, A. (2000), A measurement system for online course delivery success. Retrieved August 1,
2002, from Georgia State University, eCommerce Institute, Faculty Development Committee
Web site: http://robinson.gsu.edu/fdc/FinalReportRai2000.pdf
REVELLA, J. B, MORAN, J. W., & COX, C. A (eds.), The QFD handbook (pp. 3-12), New York:
John Wiley, 1997.
RYAN, S., SCOTT, B., FREEMAN, H. & PATEL, D. (2000), The virtual university: The Internet
and resource-based learning. Sterling, VA: Stylus.
STEIN, J. (2002), Increasing the quality and efficiency of web-based course production, Proceedings of
the 18th Annual Conference on Distance Teaching and Learning, USA.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

161
Multi-Criteria Decision Model for e-learning Architecture
Selection based on Utility Function and ELECTRE Method

Hamed Fazlollahtabar

Young Researchers Club, Islamic Azad University Babol Branch, Iran
E-mail: hamed@ustmb.ac.ir


Abstract
An e-learning architecture selection has been analyzed. This is a typical problem
when dealing with e-learning system selection. For each alternative of an e-learning
architecture there is an evaluation of both cost and quality of service. The latter may
include probabilistic delivery time and confidence in quality commitment. The
decision-maker takes into account multi-criteria evaluation through ELECTRE
method. Besides, each criterion is evaluated through a utility function. The model
integrates both approaches to indicate the best e-learning architecture. This paper
presents the formulation for the decision model and a numerical application to
illustrate the use of the model.

Keywords: e-learning Architecture Selection; Multi-criteria Decision; Utility Theory;
ELECTRE Method.


1. Introduction

E-learning refers to the use of electronic devices for learning, including the delivery
of content via electronic media such as Internet/Intranet/Extranet, audio or video tape,
satellite broadcast, interactive TV, CD-ROM, and so on (Kaplan-Leiserson, 2000). This
type of learning moves the traditional instruction paradigm to a learning paradigm
(Jönsson, 2005), thereby relinquishing much control over planning and selection to the
learners. In addition, it offers the following advantages to learners: Cost-effectiveness,
timely content, and access flexibility (Hong, Lai and Holton, 2003; Lorenzetti, 2005;
Rosenberg, 2001).
The capability and flexibility of the web-based e-learning systems (WELSs) having
been demonstrated in both training and education, resulted in their adoption by academia
as well as industry. Since the commercial application package (or commercial off-the-shelf)
strategy of system development is so widespread (Whittenet et al. 2004), the proliferation
of WELS applications has caused confusion for the potential adopters having to make
selective decisions from among candidate products or solutions. At the same time,
organizations with adopted systems are faced with issues arising from the post-adoption phase.
Several studies have been conducted on the process of e-learning dealing with
different aspects of this matter. Regarding the e-learning architecture selection little
research has been found in the literature. Almeida (2001) has dealt with maintenance
architecture selection based on a multi-criteria model, which uses contributions from
University of Bucharest and Ovidius University of Constanta


162
multi-attribute utility theory (MAUT). Dulmin and Mininno (2003) have presented a
multi-criteria decision aid method, so called PROMETHEE/GAIA to approach a e-learning
system provider’s selection model, which is applied in the context of the rail organization
employees. Multi-criteria decision aid methods such as PROMETHEE/GAIA and MAUT
allow the decision-maker to quantify multiple objectives even when these objectives
contain conflicting attributes or when they are subjective.
This paper presents results of research dealing with a multi-criteria decision model
for e-learning architecture selection, using contributions from utility theory associated
with the ELECTRE method. The problem consists of selecting the most appropriate
alternative for an e-learning architecture taking several criteria into account. Each
architecture alternative implies a specific cost and service quality characteristics.
Probabilistic delivery time, confidence commitment regarding deadlines and service
quality are among these service quality characteristics. The decision-maker has to choose
the best alternative taking into account the consequences modeled through a multi-criteria
method. Utility function (Keeney and Raiffa, 1976) and ELECTRE method (Roy, 1996)
have been taken into account to model this problem.


2. E-Learning Architecture Selection Problem

E-learning is a popular business strategy nowadays, and requires close attention to
the appropriate architecture selection process. The e-learning implementation price is no
longer the only aspect to be taken into account regarding decisions on e-learning
architecture selection. That is, different aspects have to be considered by the decision-
maker, such as cost of the architecture and performance of the service. In general, service
delivery time is a formal commitment written in the e-learning implementation contract.
Normally a different service delivery time implies a specific condition of resources,
personal skills and service availability and may result in a different cost. The architecture
is assumed to have basic variables related to multiple objectives. These multiple
objectives may be represented through the performance objectives of a production
strategy plan, such as quality, speed, dependability, flexibility, and cost (Slack et al.,
1995). Other approaches for multiple objectives may be obtained from multidimensional
quality views (Evans and Lindsay, 1989; Teboul, 1990; Garvin, 1988). The Action Space
corresponds to the set of alternatives available to the decision-maker. An action element of the
set is represented by a. The set of all actions is discrete with m elements: {a
1
, a
2
. . . , a
m
}.
Each element of this set corresponds to a possible e-learning architecture to be adopted by
the decision-maker that faces the problem.
In this paper the e-learning architecture selection problem is analyzed with respect
to the following criteria: cost, service delivery time and flexibility. Thus, for each action
a
i
there is a related cost c
i
, and specific conditions associated to the service delivery time
t
i
and its flexibility d
i
. For each action a
i
, c
i
is assumed to be a constant value and t
i
is a
random variable. So, the decision model incorporates the uncertainty associated with t
i

through the probability density function f
i
(t
i
). Flexibility is represented by the probability
d
i
for achievement of architecture conditions.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

163
3. Multi-Criteria Decision Approaches

Several multi-criteria decision methods are available (Vincke, 1992; Brans and
Mareschal, 2002; Belton and Stewart, 2002) to deal with this kind of problem. The
method should be chosen considering the nature of the problem and the model building
process. Regarding the model presented in this paper, two of these methods are briefly
described. MAUT (Keeney and Raiffa, 1976; Belton and Stewart, 2002) allows the
decision-maker to quantify and aggregate multiple objectives even when these objectives
are composed of conflicting attributes. The decision-maker’s preferences are modeled in
order to obtain a multi attribute utility function, for instance U(c
i
, t
i
, d
i
). This function
aggregates utility functions for all criteria or attributes. That is, an analytical function is
obtained which combines all criteria through a synthesis function. Each particular
analytical form for this function has preferential independence conditions to be evaluated,
in order to guarantee that the decision maker’s preferences are associated to the basic
axioms of the theory (Almedia, 2001; Keeney and Raiffa, 1976).
ELECTRE method provides a different approach. This method concentrates the
analysis on the dominance relations among the alternatives. That is, this method is based
on the study of outranking relations, exploiting notions of concordance (Roy, 1996,
Vincke, 1992; Brans and Mareschal, 2002; Belton and Stewart, 2002). These outranking
relations are built in such a way that it is possible to compare alternatives. The
information required by ELECTRE consists of information among the criteria and
information within each criterion (Roy, 1996). The method uses concordance and
discordance indexes to analyze the outranking relations among the alternatives. These
indexes are obtained through the following relations, considering two actions: a and b:

,
) (
) (
) , (


− = +
= +
+ +
+
=
W W W
W W
b a C (1)



=

k k
ak bk
Z Z
Z Z
Max b a D
*
) (
) , ( , for all k where ,
ak bk
Z Z > (2)

a S b if p b a C ≥ ) , ( and , ) , ( q b a D ≤ so called outranking relation, (3)

where, C(a, b) is the concordance index that action a outranks action b, D(a, b) is the discordance
index that action a outranks action b, a S b corresponds to the outranking relation; it means
that action a outranks b, p is the concordance index threshold, q is the discordance index threshold,
W
+
corresponds to the sum of weights for criteria where a is preferable to b, W
=
corresponds
to the sum of weights for criteria where a = b, W

corresponds to the sum of weights for criteria
where b is preferable to a, Z
ak
is the evaluation or utility of action a related to criteria k,
*
k
Z is the
best degree of evaluation obtained for criteria k and

k
Z is the worst degree of evaluation
obtained for criteria k. In order to facilitate the procedure, the evaluation of alternatives
are normalized such that 1
*
=
k
Z and 0 =

k
Z .
University of Bucharest and Ovidius University of Constanta


164
The outranking relation is obtained by applying both equation (3) and the procedure
to obtain the kernel, which is the sub-set of the best alternatives. The kernel includes a
sub-set of alternatives where any other alternative is outranked by at least one of the
kernel and the alternatives of the kernel are incomparable.


4. Decision Model

The decision-maker’s preferences for each criterion are modeled in order to obtain
the utility function for each objective of the architecture. Then, the utility function is
obtained from the decision-maker for each consequence: U(t), U(c) and U(d). These
utility functions are obtained by applying one of the classical elicitation procedures
(Berger, 1985; Raiffa, 1970). The final solution depends on the utility function for each
criterion. The analytical form of this utility function may represent one of the three basic
conditions for the decision-maker behavior. That is, aversion, neutral or prone condition
may be considered (Berger, 1985; Raiffa, 1970). For U(t) it is assumed that the decision-
maker behavior is accordingly of exponential analytical form given below:


t A
e t U
1
) (

= (4)

The exponential utility function is a typical function often found in practice (Raiffa,
1970) for one-dimensional utility functions. In previous work (Almedia, 2001) the
exponential utility function has been found for U(t) and U(c). This means that higher
values of t or c are much more undesirable for the decision-maker. Thus,


c A
e c U
2
) (

= (5)

For U(d) is assumed to be a linear function:

d A d U
3
) ( = (6)

Therefore, the evaluation of variable t is given by the decision-maker through the
utility function U(t). However, the evaluation of alternatives is based on the probabilistic
characteristics of t. Thus, a probability density function f (t) for t is taken into account.
The assumption of f (t) implies different results, given the decision-maker’s preferences
for this probabilistic criterion. Gamma probability function, with parameter n = 2, is
assumed for f (t). This condition may be found in practical situations where delivery time
is concentrated around a modal value. Thus,


ut
te u t f

=
2
) ( (7)

The 3
rd
International Conference on Virtual Learning, ICVL 2008

165
Once U(t) gives the evaluation for variable t , the evaluation of the alternatives is
based on parameter u. Then, U(u) is derived based on U(t). Applying the linearity
property of utility theory (Berger, 1985) for utility of t, it follows that




=
0
dt ) ( ) ( ) ( t f t U u U E
t
(8)

Applying (4) and (7) into (8), it follows that

| |


− −
=
0
2
dt, ) (
1
t A ut
t
e e tu u U E

2
1
2
) (
) (
u A
u
u U E
t
+
= , (9)

Therefore, in order to incorporate the probabilistic aspect related to t, this variable
is represented by its parameter u. Thus, for each action a
i
, a parameter u
i
is applied. Once
the utility function is obtained for all criteria U(u), U(d) and U(c), the ELECTRE method
may be applied. In this case, the decision-maker establishes the relative weights for the
criteria, taking into account framework of the non-compensatory ELECTRE method
(Roy, 1996; Vincke, 1992). In the MAUT approach the decision-maker preferences are
modeled in order to obtain a multi-attribute utility function U (u, c, d), when aggregates
all utility functions U(c), U(d) and U(u). The function U(u, c, d) has to be evaluated in
order to guarantee that the axioms of the theory (MAUT) conform to the decision-
maker’s preferences.
A different approach is employed by the ELECTRE method. This method exploits
some characteristics of dominance regarding the multiple criteria analyzed. In this
method a concordance notion allows the ranking of alternatives, analyzing outranking
relations among alternatives (Vincke, 1992). This allows a decision support approach
avoiding rigid assumptions required by MAUT from the decision-maker (Vincke, 1992).
The ELECTRE method is based on the study of outranking relations, using a non-
compensatory logic.
Each alternative of architecture can be evaluated through:
• Architecture cost c,
• Parameter u, associated to the probability density function of t, and
• Architecture flexibility d.
The ELECTRE method may now be applied to (5), (6) and (9) given to the
decision-maker the best alternatives of architecture.


5. Numerical Application

In order to illustrate the use of the decision model, there follows a presentation of a
numerical application. This application is based on a case regarding service outsourcing
University of Bucharest and Ovidius University of Constanta


166
related to transportation. In this context the cost and the response time t has
characteristics suitable to the exponential utility function previously discussed.
The cost is given in monetary units for each architecture alternative as follows:
1) Action a
1
– most expensive (c
1
= 100) and reduced t , such that u = 0.95; d = 0.95.
2) Action a
2
– medium cost (c
2
= 60) and medium t , such that u = 0.65; d = 0.90.
3) Action a
3
– least expensive cost (c
3
= 10) and the large t , such that u = 0.03; d = 0.75.
4) Action a
4
– below medium cost (c
4
= 30) and t , such that u = 0.15; d = 0.8.
5) Action a
5
– below medium cost (c
5
= 50) and t , such that u = 0.70; d = 0.75.
6) Action a
6
– expensive cost (c
6
= 85) and reduced t , such that u = 0.9; d = 0.8.
The following weights have been applied: for c: 0.40; for d: 0.25; t : 0.35.

Table 1
Architecture Alternatives and their Related Normalized Utilities

Architecture alternatives Normalized utilities for alternatives
Action u C d U‘(u) U‘(C) U‘(d)
a
1
0.95 100 0.95 1.00 0.00 1.00
a
2
0.65 60 0.90 0.91 0.24 0.75
a
3
0.03 10 0.75 0.00 1.00 0.00
a
4
0.15 30 0.80 0.40 0.61 0.25
a
5
0.70 50 0.75 0.93 0.34 0.00
a
6
0.90 85 0.80 0.99 0.07 0.25

E-learning architecture alternatives are given in Table 1.Values of criteria are
assigned to all alternatives a
i
, where i = 1, 2, . . . , 6. The utility function for cost is given
in (5) with the parameter A
2
= 0.02. The utility function for d is given in (6) with parameter
A
3
= 1. The utility function for t is given in (9) with parameter A
1
= 0.1. Applying these
equations (5), (6) and (9), utilities are obtained for all alternatives. Table 1 also presents
U‘(u), U‘(C) and U‘(d), corresponding to normalized values for utilities. The normalization
procedure is based on a linear transformation. For instance U‘(u) = aU(u) + b, such that
a>0. According to the utility theory (Keeney and Raiffa, 1976) this linear transformation
insures that U‘(u) is strategically equivalent to U(u). That is, U‘(u) preserves the same
properties and the preference structure of U(u). Based on decision-maker’s preferences,
the weights of criteria have been assigned as previously mentioned and the admissible
levels (thresholds) for concordance index and discordance index are as follows: p = 0.5
and q = 0.45. Therefore, when (1) and (2) are applied, the concordance and discordance
indexes are obtained. Then, the outranking relation is obtained by applying (3). Finally,
by applying the procedure to obtain the kernel (Roy, 1996; Vincke, 1992; Belton and
Stewart, 2002), alternatives a
1
and a
4
are identified. This result indicates that a
1
and a
4
,
although incomparable between themselves, are the two best alternatives for the
preferences presented by the decision-maker and the assumptions underlying the model.
A sensitivity analysis of weights and admissible levels for the concordance index and the
discordance index (varying by 10%) shows that the result remains the same. This analysis
indicates that the recommendation for action a
1
and a
4
are sufficiently robust, regarding
the limits of variation mentioned above.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

167
6. Conclusion

Two different multi-criteria approaches have been applied to deal with multiple
criteria in similar problems: MAUT and the PROMETHEE method. The development of
criteria scales to identify the intensity of preference for one alternative over another is
based on deterministic way. The problem approached in this paper is related to the
context of e-learning architecture, where uncertainties of some variables are relevant,
such as: service delivery time t
i
and flexibility d
i
, for a given alternative a
i
. Then, a utility
function is introduced in order to incorporate the uncertainty evaluation of those
variables. Theses utility functions are integrated into the ELECTRE framework in order
to obtain multi-criteria evaluation within a non-compensatory approach.
The main requirements of this theory imply a rationality that involves
compensation among the criteria, which involves the procedure for aggregation of all
criteria obtaining a synthesis multi-criterion utility function. This rationality is not always
accepted by the decision-maker. The decision-maker rationality may require a non-
compensatory method, where the decision support process does not require an
aggregation of all criteria. The ELECTRE method may support decision process under
this situation. The model proposed in this paper presents an alternative approach for
analyzing an e-learning architecture selection problem. Using utility theory each criterion
is represented by a utility function, incorporating the probabilistic structure of the
problem. The probability function for the service delivery time is assumed to be gamma
probability density function. The evaluation of the criteria represented by each utility
function is analyzed through the ELECTRE method. The paper includes the structure of
the decision model to support the decision-maker and a numerical application illustrates
the use of the model.


7. Acknowledgments

This research has been supported by Mazandaran Telecommunication Research Center.



REFERENCES


ALMEIDA, A. T., Multicriteria decision making on maintenance: spares and contracts planning.
European Journal of Operational Research 2001, 129(2):235-41.
BELTON, V., STEWART, T. J., Multiple Criteria Decision Analysis, Dordrecht: Kluwer, 2002.
BERGER, J. O., Statistical Decision Theory and Bayesian Analysis, Berlin: Springer, 1985.
BRANS, J. P., MARESCHAL, B., PROMةTةE-GAIA – une méthodologie d’aide à la décision em
présence de critères multiples, Brussels: Editions de l’Université de Bruxelles, 2002.
DULMIN, R, MININNO, V., Supplier selection using a multi-criteria decision aid method. Journal
of Purchasing and Supply Management 2003, 9: 177-87.
EVANS, J. R., LINDSAY, W. M., The Management and Quality Control,West Publishing
Company, 1989.
University of Bucharest and Ovidius University of Constanta


168
GARVIN, D. A., Managing Quality the Strategy and Competitive Edge, NewYork: Free, 1988.
HONG, K. S., LAI, K. W. and HOLTON, D. (2003), Students’ satisfaction and perceived learning
with web-based course, Educational Technology and Society, 6(1), 116-124.
JÖNSSON, B.-A. (2005), A case study of successful e-learning: a web-based distance course in
medical physics held for school teachers of the upper secondary level, Medical Engineering
and Physics, 27(7), 571-581.
KAPLAN-LEISERSON, E. (2000), E-Learning glossary. Available from http://www.learningcircuits.org
KEENEY, R. L., RAIFFA, H., Decision with multiple objectives: preferences and value trade-offs,
NewYork:Wiley, 1976.
LORENZETTI, J. P. (2005), How e-learning is changing higher education: A new look. Distance
Education Report, 22 (July), 4-7.
RAIFFA, H., Decision Analysis, Reading, MA: Addison-Wesley, 1970.
ROSENBERG, M. J. (2001), E-learning: Strategies for delivery knowledge in the digital age, New
York, McGraw-Hill.
ROY, B., Multicriteria for decision aiding, London: Kluwer, 1996.
SLACK, N., CHAMBERS, S., HARLAND, C., HARRISON, A., JOHNSON, R.. Operations
Management, London: Pitman Publishing, 1995.
TEBOUL, J., La Dynamique qualité, Paris: Les Editions d’Organisation, 1990.
VINCKE, P., Multicriteria Decision-aid, NewYork:Wiley; 1992.
WHITTEN, J. L., BENTLEY, L. D. and DITTMAN, K. C. (2004), System Analysis and Design
Methods, New York, McGraw-Hill.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

169
Applying Integrated Strategic Planning and RADAR Technique
to Find Optimal Course Delivery Policy
in a Virtual Learning System

Hamed Fazlollahtabar
1
*, Ali Abbasi
2

1
Young Researchers Club, Islamic Azad University Babol Branch, Iran
2
Department of Industrial Engineering,
Mazandaran University of Science and Technology, Babol, Iran
*e-mail: hamed@ustmb.ac.ir


Abstract
The popularity of the Internet as an information source has grown extensively. Its
shear expanse and convenience is ideal to disperse information. More and more
online services have now become available such as online banking, e-government, e-
learning and e-commerce. Our interest lies with e-learning and in particular with the
delivery of course material online. Strategic management can be understood as the
collection of decisions and actions taken by business management, in consultation
with all levels within the organization, to determine the long-term activities of the
organization. Many approaches and techniques can be used to analyze strategic
cases in the strategic management process. Among them, Strengths, Weaknesses,
Opportunities and Threats (SWOT) analysis, which evaluates the opportunities,
threat s, strengths and weaknesses of an organization, is the most common. SWOT
analysis is an important support tool for decision-making, and is commonly used as
a means to systematically analyze an organization’s internal and external
environments. In this paper we apply SWOT analysis to evaluate possible strategies
to deliver an online course in a virtual learning environment. Then using RADAR
technique we rank the strategies and select the optimal one.


1. Introduction

The popularity of the Internet as an information source has grown extensively. Its
shear expanse and convenience is ideal to disperse information. More and more online
services have now become available such as online banking, e-government, e-learning
and e-commerce. Our interest lies with e-learning and in particular with the delivery of
course material online. More specifically, we are interested in presenting online course
material in interactive and stimulating ways for students and creating an online learning
community similar to that which one might experience in an actual university. In this
article, we present our experience of developing an innovative collaborative e-learning
system. As technologies have advanced, so too have the delivery methods for e-learning.
Early forms included CDROMs and knowledge pools on the Internet, where users could
access information and work through it at their own pace. This has now progressed to
course and learning management systems, which provide greater support to tutors and
University of Bucharest and Ovidius University of Constanta


170
students. Learning Management Systems (LMSs) which are now available provide course
administration tools for instructors, allowing them to manage the distribution of course
material and assignments. The importance of communication and collaboration within e-
learning has been highlighted previously by Preece (2000); Hamburg, Lindecke, and ten
Thij (2003); Salmon (2002) and Thurmond and Wambach (2004) amongst others, and as
a result online forums and discussion boards have become an invaluable resource in these
LMSs. They allow students to communicate with their peers and tutors thus empowering
them to socialize and learn together online. While e-learning systems have improved with
time, we feel that there are still some issues to be resolved before a truly stimulating and
realistic learning experience can be provided online. Partaking in an online course can be
a much more engaging and interactive experience for students.
Through the use of technologies such as Virtual Reality (VR) and instant communication,
students can be more visually aware of their classmates and can converse in real-time
with them. They can also receive immediate feedback from their tutors and gain a sense
of being present in the same place as their peers despite their remote physical locations.
These shared virtual environments also facilitate simultaneous viewing of learning
materials by the whole class and allow them to actively partake in group discussions
about the learning content at the same time.
VR has been very popular and successful in other areas including entertainment and
urban planning. It has also been extensively used within manufacturing industries and
military bodies (Burdea & Coiffet, 2003). In addition, the benefits of 3D graphics for
education have been explored. Many 3D resources have already been developed in this
area. 3D models are very useful to familiarize students with features of different shapes
and objects, and can be particularly useful in teaching younger students. Many games
have been developed using 3D images that the user must interact with in order to learn a
certain lesson. Interactive models increase a user’s interest and make learning more fun.
3D animations can be used to teach students different procedures and mechanisms for
carrying out specific tasks (Nijholt, 2000; Rickel & Johnson, 1999). VR has also been
used extensively for simulations and visualization of complex data. For example, medical
disciplines use VR to represent complex structures (Ryan, O’Sullivan, Bell, & Mooney,
2004) and increasingly scientists are using this technology for visualization and in
particular as a teaching aid (Manseur, 2005).
The use of VR and 3D graphics for e-learning is now being further extended by the
provision of entire VR environments where learning takes place. This highlights a shift in
e-learning from the conventional text-based online learning environment to a more
immersive and intuitive one. Since VR is a computer simulation of a natural environment,
interaction with a 3D model is more natural than browsing through 2D web pages looking
for information. These VR environments can support multiple users, further promoting
the notion of collaborative learning where students learn together and often from each
other (Kitchen & McDougall, 1998).
As with a real university, students are aware of each other within the environment
and they can partake in lectures, group meetings and informal chats. We feel that social
interaction is vitally important within any learning scenario and so we provide many
communication facilities in addition to learning content. VR can bring a great deal to an
e-learning experience in these ways and in this article we discuss our techniques in detail.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

171
While we recognize the importance of pedagogy in any learning scenario, pedagogic
issues relating to learning strategies and learning content are not dealt with in this article.
Instead we focus on the design and usability of a 3D interface for learning, socializing and
communicating online, and on providing adequate support for a variety of learning tasks.
In this paper we initially will analyze course delivery in strategic view point via
SWOT factors. Then we measure each policy using RADAR technique to find the
optimal one. Next section provides a comprehensive description of SWOT factors.


2. SWOT Factors

Strategic management can be understood as the collection of decisions and actions
taken by business management, in consultation with all levels within the organization, to
determine the long-term activities of the organization (Houben et al., 1999). Many
approaches and techniques can be used to analyze strategic cases in the strategic
management process (Dincer, 2004). Among them, Strengths, Weaknesses, Opportunities
and Threats (SWOT) analysis, which evaluates the opportunities, threat s, strengths and
weaknesses of an organization, is the most common (Hill and Westbrook, 1997). SWOT
analysis is an important support tool for decision-making, and is commonly used as a
means to systematically analyze an organization’s internal and external environments
(Kurttila et al., 2000; Stewart et al., 2002). By identifying its strengths, weaknesses,
opportunities, and threats, the organization can build strategies upon its strengths,
eliminate its weaknesses, and exploit its opportunities or use them to counter the threats.
The strengths and weaknesses are identified by an internal environment appraisal while
the opportunities and threats are identified by an external environment appraisal (Dyson,
2004). The internal appraisal examines all aspects of the organization covering, for
example, personnel, facilities, location, products and services, in order to identify the
organizations strengths and weaknesses. The external appraisal scans the political,
economic, social, technological and competitive environment with a view to identifying
opportunities and threats. The environmental SWOT analysis is indicated in Figure 1.

SWOT Matrix
Weakness Strength Opportunity Threat
Environmental Appraisal
External Environment Appraisal Internal Environment Appraisal


Figure 1. Environmental SWOT Analysis

SWOT analysis summarizes the most important internal and external factors that
may affect the organization’s future, which are referred to as strategic factors (Kangas et
University of Bucharest and Ovidius University of Constanta


172
al., 2003). The external and internal environments consist of variables which are outside
and inside the organization, respectively. The organization’s management has no short-
term effect on either type of variable. Comprehensive environmental analysis is important
in recognition of the variety of internal and external forces with which an organization is
confronted. On the one hand these forces may comprise potential stimulants, and on the
other hand, they may consist of potential limitations regarding the performance of the
organization or the objectives that the organization wishes to achieve (Houben et al,
1999). The obtained information can be systematically represented in a matrix (Ulgen and
Mirze, 2004); different combinations of the four factors from the matrix can aid in
determination of strategies for long-term progress. When used properly, SWOT can
provide a good basis for strategy formulation (Kajanus et al., 2004). However, SWOT
analysis is not without weaknesses in the measurement and evaluation steps (McDonald,
1993). In conventional SWOT analysis, the magnitude of the factors is not quantified to
determine the effect of each fact or on the proposed plan or strategy (Masozera et al.,
2006). In other words, SWOT analysis does not provide an analytical means to determine
the relative importance of the factors, or the ability to assess the appropriateness of
decision alternatives based on these factors. While it does pinpoint the factors in the
analysis, individual factors are usually described briefly and very generally. More
specifically, SWOT allows analysts to categorize factors as being internal (Strengths,
Weaknesses) or external (Opportunities, Threats) in relation to a given decision, and thus
enables them to compare opportunities and threats with strengths and weaknesses
(Shrestha et al., 2004). Based on the aforementioned description the following SWOT
matrix (Table 1) is configured for online course delivery in an e-learning system.

Table 1
SWOT matrix

Strengths:
S
1
: The most up-to-date facilities
S
2
: The well educated employees
S
3
: Rapid delivery
S
4
: The strong link between e-learning
center and user
Weaknesses:
W
1
: Concentration of industries in specific
points
W
2
: Using modern equipments which
would incur more costs
W
3
: Loss of relationship between user and
teacher
Opportunities:
O
1
: Paying attention to modern
technologies
O
2
: Special attention to software revolution
O
3
: Communication growth with lower
costs
Threats:
T
1
: Laziness of students
T
2
: Health danger of high tech equipments


Based on the above SWOT matrix the following strategies could be proposed to
develop online course in an e-learning system:
1. Establishing an integrated information system to facilitate data transfer among
users. (S
3
, S
4
, W
1
, O
2
, O
3
).
The 3
rd
International Conference on Virtual Learning, ICVL 2008

173
2. The consolidated link between e-learning centers, users, and teachers. (S
2
, W
3
,
O
1
, O
4
).
3. Entering clean modern information technologies amongst users. (S
1
, W
2
, T
1
, T
2
).
4. Investing on educating employees with modern methods. (S
2
, W
3
, T
1
).
Here we identified the policies regarding to SWOT factors. Next section gives
analytical method for quantifying and ranking the strategies.


3. RADAR Technique

RADAR is an acronym of: results, approach, deployment, assessment, review. It is
used to describe a consistent and systematic approach to improvements in a given
organization, which is to be applied as a standard. The RADAR Method describes a plan-
do-check-act cycle, thus corresponding to the idea of continuous improvement, which is
inherent to the cycle.
First of all, using the RADAR Method we must define what is to be achieved from
an operational point of view (results). In the phase that follows we plan the approaches
required to achieve the results defined (approach). The planning phase is followed by a
systematic and full implementation of what has been planned (deployment). As a last
step, we monitor approaches and implementation, reviewing and assessing them together
with the results achieved (assessment and review). Thus, the use of the RADAR Method
enables us to identify the strengths of an organization and its potential for improvement.
A well-founded and systematic approach is rated positively. The quality status of a given
organization is quantified by a scale of points. The cycle described above initiates a
learning process, in which a given organization can identify and permanently improve the
causal relationships between goals, approaches and implementation.
Questionnaire: A questionnaire is used to collect information on all important
aspects and criteria. Answering options may be restricted to yes-no answers or be of the
multiple choice type to identify an organization's current quality status. The advantage of
this method is that it can be used quickly and easily. Questionnaires can be completed
even by those users who have little know-how on quality management and the RADAR
Method. The disadvantage, however, is that an organization's strengths and improvement
potentials are not explicitly identified and that the results of a self-assessment exercise are
strongly dependent on the interviewed person's awareness of problems and on the quality
of questions asked.
The result that is aimed to gain is to find the optimal strategy of online course
delivery in an e-learning system. According to last section, the approach and deployment
phases have been accomplished through the definition of different strategies. Therefore,
regarding to RADAR technique we can quantify the policies for assessment and review
phases. Some sub-criteria could be defined for RADAR. The sub-criteria and the
quantification process is done by the following conventional table (Table 2).

University of Bucharest and Ovidius University of Constanta


174
Table 2
Numerical Value Associated with Description for Ranking

Score

RADAR
1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Results Trends Targets Comparisons Causes Scope
Approach Sound Integrated – – –
Deployment Implemented Systematic – – –
Assessment
&
Review
Measurement Learning Improvements – –

For each of the RADAR criteria (Result, Approach, Deployment, Assessment and
Review) we have some sub-criteria. A decision maker can judge and give a numerical
value from 1 to 5 (where 1 is the worst and 5 is the best). Then by summing up the
numerical value of sub-criteria for Result, we gain the value of Result. The same thing
should be done for Approach, Deployment, Assessment and Review. After that a
summation of Result, Approach, Deployment, Assessment and Review numerical values
should be achieved. The result is total score for each policy. Consequently the results of
all policies are compared and the one with the highest RADAR score is selected as
optimal policy of online course delivery in an e-learning system of education.


4. Conclusions

Strategic planning is a significant element for implementing projects. A benefit
method for analyzing different strategies is SWOT. SWOT analysis summarizes the most
important internal and external factors that may affect the organization’s future, which are
referred to as strategic factors. The RADAR Method describes a plan-do-check-act cycle,
thus corresponding to the idea of continuous improvement, which is inherent to the cycle.
In this paper, integration between SWOT analysis and RADAR technique is applied to
evaluate varied strategies for selecting optimal online course delivery in an e-learning
system. The most important advantage of the proposed approach is including qualitative
assessment in implementing policies. That qualitative analysis helps the organization in
continuous improvement.


5. Acknowledgment

This research has been supported by Mazandaran Telecommunication Research Center.



The 3
rd
International Conference on Virtual Learning, ICVL 2008

175
REFERENCES


BURDEA, G. C., & COIFFET, P. (2003), Virtual reality technology (2nd ed.), John Wiley & Sons.
G. HOUBEN, K. LENIE, K. VANHOOF, A knowledge-based SWOT analysis system as an
instrument for strategic planning in small and medium sized enterprises, Decision Support
Systems 26 (1999), 125-135.
H. ULGEN, S. K. MIRZE, Strategic Management, Literatur Publication, Istanbul, 2004.
HAMBURG, I., LINDECKE, C., & TEN THIJ, H. (2003), Social aspects of e-learning and blending
learning methods, in Proceedings of the fourth European conference on E-commerce, E-work,
E-learning, E-health, E-banking, E-business, on-line services, virtual institutes, and their
influences on the economic and social environment (E-Comm-Line), pp. 11-15.
J. KANGAS, M. KURTILA, M. KAJANUS, A. KANGAS, Evaluating the management strategies
of a forestland estate-the S-O-S approach, Journal of Environmental Management 69
(2003), 349-358.
J. TERRADOS , G . ALMONACID, L. HONTORIA, Research Group IDEA, Polytechnics School,
Campus Las Lagunillas, Edificio A3, University of Jae ´n, 23071 Jae ´n, Spain , 24 August 2005.
KITCHEN, D., & MCDOUGALL, D. (1998), Collaborative learning on the Internet, Journal of
Educational Technology Systems, 27, 245-258.
M. KAJANUS, J. KANGAS, M. KURTTILA, The use of value focused thinking and the A’WOT
hybrid method in tourism management, Tourism Management 25 (2004) 499-506.
M. KURTTILA, M. PESONEN, J. KANGAS, M. KAJANUS, Utilizing the analytic hierarchy process (AHP),
in SWOT analysis-a hybrid method and its application to a forest-certification case, Forest
Policy and Economics 1 (2000), 41-52.
M. H. B. MCDONALD, The Marketing Planner, Butter-worth-Heinemann, Oxford, 1993.
M. K. MASOZERA, J. R. R. ALAVALAPATI, S. K. JACOBSON, R. K. SHRESTA, Assessing
the suitability of community-based management for the Nyungwe Forest Reserve, Rwanda,
Forest Policy and Economics 8 (2006), 206-216.
MANSEUR, R. (2005), Virtual reality in science and engineering education, in Proceedings of the
frontiers in education conference.
NIJHOLT, A. (2000), Agent-supported cooperative learning environments, in Proceedings of the
international workshop on advanced learning technologies.
O. DINCER, Strategy Management and Organization Policy, Beta Publication, Istanbul, 2004.
PREECE, J. (2000), Online communities: designing usability, supporting sociability, Chichester,
UK: John Wiley & Sons.
R. STEWART, S. MOAMED, R. DAET, Strategic implementation o f IT/IS projects in
construction: a case study, Automation in Construction 11 (2002), 681-694.
R. G. DYSON, Strategic development and SWOT analysis at the University of Warwick,
European Journal of Operational Research 152 (2004), 631-640.
R. K. SHRESTHA, J. R. R. ALAVALAPATI, R. S. KALMBACHER, Exploring the potential for
silvopasture adoption in South-central Florida: An application o f S WOT-AHP method,
Agricultural Systems 81 (2004), 185-199.
RICKEL, J., & JOHNSON, W. L. (1999), Virtual humans for team training in virtual reality, in Proceedings
of the ninth international conference on artificial intelligence in education, pp. 578-585.
RYAN, J., O’SULLIVAN, C., BELL, C., & MOONEY, R. (2004), A virtual reality
electrocardiography teaching tool, in Proceedings of the second international conference in
biomedical engineering, pp. 250-253.
University of Bucharest and Ovidius University of Constanta


176
SALMON, G. (2002), E-activities: the key to active online learning, London: Kogan Page.
T. HILL, R. WESTBROOK, SWOT analysis: it’s time for a product recall, Long Range Planning
30 (1997), 46-52.
THURMOND, V. A., & WAMBACH, K. (2004), Understanding interactions in distance
education: a review of the literature, Journal of Instructional Technology and Distance
Learning, 1, 9-33.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

177
“Evalution Traps”: A Brief Vademecum to Avoid the Most
Common Mistakes in Distance Learning Evaluation

Gaetano Bruno Ronsivalle
1
, Piera Vivolo
2


(1) Professor of University of Rome, La Sapienza
R&D Manager AbiFormazione, Milan-Rome Italy
R&D Manager LabelFormazione, Rome Italy
E-mail: sprsricercaesviluppo@abiformazione.it
(2) Researcher, LabelFormazione, Rome Italy
80, Scorticabove st, Rome, IT – 00156 – ITALY
E-mail: pvivolo@abiformazione.it


Abstract
The purpose of this paper is to draw attention on the methodological inaccuracy
often involving the macro-design of evaluation systems for e-learning courses. In
our experiences customers often prefer delivering products in record time, to the
detriment of appropriate instructional and evaluation design. For example, they ask
for web based courses with only summative tests, or for evaluation system design
without any content expert’s aid; they also often deny the time needed to verify the
validity and reliability of the designed instruments. This means that customers
underestimate the importance of evaluation by regarding it as something apart from
the training program; nevertheless, they always ask for final tests capable of
measuring the participants’ acquisition of skills and competences. It is, then,
necessary to underline the practical problems concerning with the uncontrolled
setting-up of evaluation systems from both instructional and methodological point of
view. In this paper we propose all the steps of a proper evaluation macro-design
activity and then offer some case-studies in order to point out all the problems
resulting from omissions occurred in this phase. We describe, for example, what can
happen when it is not clear: “what” to measure and “how” to do it, how to design
the sampling and to store information to easily build items or how to run empirical
analysis. Our final purpose is to create a vademecum to avoid the most common
evaluating mistakes and to make customers aware of the practical involvements of a
rigorous evaluation system.

Keywords: Assessment, evaluation design, methodology, vademecum.


1. The complex concept of evaluation

In learning context, evaluation is a very useful, but multifaceted and complex,
instrument. Very few people, among teachers, managers and insiders, are conscious of the
traps that evaluation can set. In fact, the more the interest on assessment grows up, the
more methodological inaccuracy spreads. Planning and building an evaluation system is
not so easy as it comes.
Before going on discussing the many facets of evaluation concept and all its
concerning pitfalls, it is useful to briefly analyze the phases of the evaluation process.
Our evaluation system design model consists of 5 main phases:
University of Bucharest and Ovidius University of Constanta


178
• Phase 1: Design of the tree of objectives (cf. Mager, 1975), through the matching
between learning objectives and item typologies, on the basis of Bloom’s Taxonomy
(Bloom, 1956) or other cognitive taxonomies;
• Phase 2: Analysis of the “Management variables”, that answers to questions such
as why, how and when evaluate, and also concerns data selection, report forms and
technological restrictions;
• Phase 3: Storyboard and items building up, regarding the choice and the design
of the proper item typology relating to the specific objective;
• Phase 4: Scoring models, concerning also weighted systems and rules;
• Phase 5: Review and control of the Beta version test.
Focusing on distance learning, evaluation plays an unquestionable role to guarantee
quality processes in learning contexts, for its own features and involvements (Ronsivalle,
et al. 2007). In spite of this, in our experiences, customers are not usually aware of the
importance and significance of a well-done evaluation system: they do not care about its
“methodological accuracy” and often ask for assessment only because they need
certificates of attendance or because of the general trend.
Unfortunately the problem concerns some instructional designers, too: because of
the lack of a theoretical design model, sometimes they design incorrect and inappropriate
evaluation systems.
The topic is very tough and ticklish. In the following pages we will discuss some
case-studies and show the most common evaluation “traps” at macro and micro design level.


2. Common evaluation traps: a vademecum

I. The ill-described objective trap

Example: A food industry requested the design of the evaluation system of a
distance training course on “Meat export Legislation”. During the preliminary analysis,
the general objective of the course was not properly focused. Furthermore, the customer
asked for a brief testing in order to not tire the end users who were unskilled in pc and
new technologies using.
Trouble: Items amount was not sufficient nor representative of learning contents,
so the test could not measure all the significant objectives.
Mistake: Unsuccessful identification of general learning objective and sub
objectives. The sub objectives do not reproduce the general objective cognitive difficulty.
Effects:
− False positive risk: the training on a particular topic and the attainment of certain
competences is certified, but the test measures something else.
− Damage risk for external people and Corporate image: in certain working
contexts the lack of competences can be dangerous for the collective safety.
− Money risk: loss of money put in the training intervention.


II. The protean complexity level trap (or Proteus' trap)

Example: A transport industry requested the design of the evaluation system of a
distance training course on “Safety and health at work”. The final users of the
intervention were directors, managers and workers. During the macro-design phase, the
The 3
rd
International Conference on Virtual Learning, ICVL 2008

179
different levels of responsibility concerning the diverse corporate roles and the different
levels of their expected final achievements were not highlighted properly.
Trouble: The achievement test is marked on a particular cognitive level that is
suitable for only one kind of user. Items result too much superficial or too much in-depth.
Mistake: Lack of adequate identification of items capable of measuring a peculiar
learning objective achievement for a specific user.
Effects:
− False positive risk: possibility to certify the training on a particular topic and the
attainment of certain competences at a definite cognitive level, whereas the acquired level
of achievement is higher or lower.
− Damage risk for external people and Corporate image: attainment of
competences at a level that is lower than the expected one.
− Money risk: loss of money put in the training intervention with the consequent
need of entire re-design of the learning path.


III. The missing goal trap (or Filottete's trap)

Example: A marketing society requested the design of the evaluation system of a
distance training course on “Commercial telephone call management”. During
preliminary analysis, the scope of the testing, that should have been formative instead of
summative, was not specified.
Trouble: The test does not allow to highlight problems or improvements gained so far.
Immediate feedbacks are not provided and users cannot repeat the test.
Mistake: Lack of the didactic scope declaration and of the general testing structuring.
Effects:
− Testing uselessness: Items result not satisfactory to the formative didactic scope
that should focus in-depth only upon some contents.
− Money and time risk: loss of money put in the training intervention and waste of time.


IV. The missing time esteem (Cronus’ trap)

Example: An IT society requested the design of the evaluation system of a distance
training course on “European Computer Driving Licence”. During preliminary analysis,
any testing time limit were not established by the customer.
Trouble: Platform registered very different average response times among students
that ranged from 1’45 seconds to 10 seconds. During the test, users consulted didactic
material, even if it was not allowed, and, as a consequence, lengthened their response times.
Mistake: Unsuccessful estimation of average duration of test administration and
little clarity upon the opportunity of documents consultation.
Effects:
− False positive risk: possibility to certify the training on a particular topic and the
attainment of certain competences that were not acquired and consequent corporate image damage.
− Invalid test risk and uselessness of initial and final test data comparison.
− Loss of money put in the training intervention and waste of time.
− Need of re-design the training path considering temporal and technological restrictions.


University of Bucharest and Ovidius University of Constanta


180
V. The missing selection data entry trap (or Hermes’ trap)

Example: A Landing Institution requested the design of the evaluation system of a
distance training course on “English Language”. Customer asked for final scores reports,
without highlighting any particular need to store them or to re-use them in future.
Trouble: The users’ behaviour on the single item could not be detected. Initial and
final administrations reports included only total scores and the correction grid. In spite of
platform potentiality, the registered data resulted insufficient.
Mistake: Lack of evaluation and selection of the data needed to be registered on
the platform.
Effects:
− Loss of money put in the training intervention and waste of time.
− Registered data that are not suitable for the setting scope.
− Difficult interpretation of results.
− Impossible data storage and reusability.


VI. The defective monitoring trap (or Argo’ trap)

Example: A marketing society requested the design of the evaluation system of a
distance training course on “Communication strategies”. During preliminary meetings
with the customers, the detailed characteristics of monitoring system and data needed to
be checked were not defined.
Trouble: The LMS platform did not allow to monitor all the concerning data. In particular,
it could not manage the weighted questions and the definition of the various interactions.
Mistake: Lack of analysis of the technological restrictions of the monitoring
system. Missing information for the system developer.
Effects:
− Loss of money put in the evaluation system design.
− Unsuccessful management of testing administrations and of data analysis.
− Unsuccessful management of monitoring system.


VII. The defective Reporting trap

Example: An insurance society requested the design of the evaluation system of a
distance training course on “Problem Solving techniques”. During preliminary meetings
with the customers, it was not clear that on the basis of data reports and efficacy index, a
rewarding system for the best achievers would have been put into operation.
Trouble: Data reports resulted insufficient to calculate the Efficacy Index and, so,
to define best achievers. The report form was inadequate and full of contradictions. The
society decided to draw five users among the High Scorers at final test.
Mistake: Lack of clear and detailed reports.
Effects:
− Lot of telephone calls to tutor and protests.
− Corporate image damage risk.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

181
VIII. “The one who think before” trap (or Prometeus’ trap)

Example: A telecommunications industry requested the design of the evaluation
system of a distance training course on “IT safety”. During preliminary meetings with the
customers, it was not clear that they needed an evaluation system with weighted scores
and penalties for mistakes occurring in some crucial items.
Trouble: Customer needed a weighted evaluation system as soon as possible. On
the basis of this request, the computer programmer decided to manage the scoring by himself.
Mistake: Lack of clear instructions for the computer programmer about the
assignment of weighted scorers and the general scoring rules.
Effects:
− Loss of money put in the platform design and programming.
− Assignment of the same score to items of different relevance/difficulty.
− Collection of data that represents only the mere sum of raw scores.
− Corporate image damages.


IX. The “Ambiguous response options” trap (or the Sphinx’ trap)

Example: An Italian bank requested the design of the evaluation system of a
distance training course, on commercial topics, aiming at selecting training managers for
some new agencies. The customer had little time and did not succeeded in selecting an
expert who could validate our items from a content point of view. We had to set up the
initial and final tests only on the basis of the materials and contents the bank gave us.
Trouble: Users contested the formulation of a multiple choice item. Its three wrong
answer options seemed to be very ambiguous and looked all correct.
Mistake: Lack of content validity with consequent ambiguity of the items.
Effects:
− Lack of satisfactory sampling of contents that the test should measure.
− False positive risk: possibility to certify the training on a particular topic and the
attainment of certain competences at a definite cognitive level, whereas the acquired level
of achievement is higher or lower.
− Loss of money and waste of time for the re-design of some intervention phases.

X. The “Instructions by heart” trap (or Mnemosyne’s trap)

Example: A food industry requested the design of the evaluation system of a distance
training course on “Spanish Language” aiming at the selection of some people to be
transferred in Latin America for one year. In order to measure all the cognitive levels of
knowledge achievement of language (speaking, understanding, writing and pronunciation),
we proposed different kind of items (i.e. multiple choice, association, listen & find, etc.).
Furthermore, the customer asked for limiting the time of response for each item.
Trouble: The instructions for answering the questions were provided only at the
beginning of the test. Not recalling all the information, users could quickly solve only
multiple choice items. Incapable of giving the right answers to the remaining items, in the
established time interval, users asked to nullify the exam.
Mistake: Lack of instructions to solve the questionnaire in the right way.
Effects:
− Lack of validity and reliability of testing.
University of Bucharest and Ovidius University of Constanta


182
− Loss of money and waste of time because of questionnaire re-design.
− Corporate image damages.


XI. The missing clues trap (or Daedalus’ trap)

Example: A services company requested to revise the macro-design of its evaluation
system of a distance training course on “Door to door selling techniques ” in order to
select people to engage in this activity. This selection system, in fact, resulted inadequate
to the scope.
Trouble: The most candidates failed whether in the diagnostic test, or in the final test.
Mistake: Lack of a well-designed storyboard with guide lines to design and formulate
the items (related didactic objective, cognitive complexity level and item typology).
Effects:
− False Negative risk: possibility to certify the lack of training on a particular topic
and of the attainment of certain competences.
− Inability to discriminate in the class, the students who really acquired expected
competences and those who had not.
− Risk of selecting persons who do not fit the role, or of missing others who do it.
− Loss of money put in the evaluation/selection system.


XII. The novel micro-designer trap (or Hebe’s trap)

Example: An e-learning society requested to revise the micro-design of their
evaluation system of a distance training course on “Safety at construction site”. During
the first meeting with the customer, we ascertained that both the training course and the
evaluation system were well-designed. In spite of this, the items micro-design caused
some problems.
Trouble: During the validation phase, the subjects of the random representative
sample succeeded in answering well to all questions, both in initial and final tests items.
Mistake: Lack of clarity in the setting up of the questions, and, above all, of the
answer options that do not result fair attractive.
Effects: Loss of money and waste of time because of the needed re-formulation of
the items.


XIII. The Item-analysis underestimation trap

Example: A debt collection society requested the design of an evaluation system of
a distance training course on “Aggressiveness Management”. In spite of the interest
showed for all the designing proposals, the customer denied to fund item analysis
activities because he considered them pointless.
Trouble: The initial and final test means do not differ significantly as expected.
Both tests result too much difficult.
Mistake: Lack of pre-test and post-test item analysis in order to detect any
troublesome item and remove or modify it before testing the users.
Effects:
− Lack of validity and reliability of both tests.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

183
− Lack of possibility to detect the level of each user’s achievement with certainty.
− False positive risk: possibility to certify the training on a particular topic and the
attainment of certain competences that were not acquired.
− Loss of money, more than the investment required for item analysis.


XIV. The Omitted comparison trap (or Narcissus’ trap)

Example: An insurance society requested the design of the evaluation system of a
distance training course on a new insurance policy. During the first meeting with the
customer, his main interest came out: he was, in fact, more interested in certifying that the
training had taken place, rather than really training his agents. He asked only for a
summative test. After some months, the customer asked data to justify money investment
on the training course. Data were, obviously, insufficient for that.
Trouble: It was not possible to compare scores obtained by the users at the
beginning and at the end of the intervention.
Mistake: Lack of design of a diagnostic test to detect previous knowledge and to
compare with the final test in order to measure the learning Efficacy Index.
Effects:
− Uncertainty of the training path quality.
− Risk of difficult quantification of the amount of added learning value of the course.
− Usefulness of data for methodological analysis.
− Risk to annoy skilled users, by not providing them learning individual paths on
the basis of their diagnostic results.

XV. The fake benefit trap

Example: A services company requested to revise the evaluation system design of
a distance training course on “Personal data management Legislation”. Despite the
design of an initial and a final test in order to compare scores, the analysis of their system
highlighted an absolute absence of methodology in the design of the whole training path.
In fact, calculated on their data, the Efficacy Index resulted surprisingly negative.
Trouble: In spite of improving the knowledge system of the users, the course
created only confusion and bewilderment.
Mistake: Use of a mistaken correction key that causes the unbalancing scores or, if
the worst comes to the worst, absence of intervention and evaluation macro-design.
Effects:
− Uncertainty of the training path quality.
− Risk of difficult quantification of the learning efficacy and amount of added
learning value by the course.
− Risk of unjustified money investment.


3. Conclusion

In our opinion, in a methodological design approach to evaluation, the common
sense is not sufficient. On the contrary, it represents the main trap among the above
mentioned ones.
University of Bucharest and Ovidius University of Constanta


184
At the same time, we and all insiders know that errors can occur also with a solid
design model at the basis.
With this brief vademecum about the most common mistakes, we hope to have
highlighted the variables that, according to us, contribute to make a good job and to avoid
evaluation traps.
We are going to continue our empirical research in order to investigate solutions
and prevent the catastrophic consequences of a mistaken evaluation design.


4. Acknowledgments

A special thanks to dr.ssa Simona Carta for her revision of the present essay.



REFERENCES


Books

BLOOM, B. S. (1956), Taxonomy of Educational Objectives: The Cognitive Domain, New York: David
McKay Co Inc.,.
MAGER, R. (1975), Preparing Instructional Objectives, Belmont, CA: Lake Publishing Co.
ROMISZOWSKI, A. J. (1999), Designing Instructional Systems, 7th edition, London.


Conference Proceedings

RONSIVALLE, G. B, (2006), “A model for Quality Assessment, Measurement and Evaluation of E-Learning
in the Italian Banking Sector”, in Proceedings of The Fifth International Internet Education Conference
and Exhibition, Cairo, Egypt.
RONSIVALLE, G. B, CARTA, S., VIVOLO, P. (2007), “New quantitative models and tools for Quality
Management in Banking Educational Institutions and Departments”, in Proceedings of The Fourth
Annual Conference of Learning International Networks Consortium, Amman, Jordan.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

185
Specifications of the "Informatisation" Processes
for Productive-Instructive Workflows

Ioan Rosca
1
, Val Rosca
2


(1) LICEF institute- Teleuniversity of Montreal
E-mail: ioan.rosca@licef.teluq.uqam.ca
(2) Amazon development center - Iasi
E-mail: rosca@amazon.com


Abstract
Interlacing to do and to know justifies flexible support of instructive-productive
workflows. To assist beneficiary systems B can consist in supply, reorganization or
instruction. Support resources Rb are conceived- for emergent activities. Complex
workflows can be orchestrated through Mbo models of B procedures, activated as
"functions". To produce Rb resources, a P development procedure can base the
"requirements engineering" on Mb models- seen as "use cases". The P workflow
can be centred on the progressive concretisation of Mb as functions - that first
become Mbt testbeds and then Mbs support instruments (or Mbi instruction tools).
Such sequences (and any software production workflows) can be modelled as Mp
"functions", acting as development coordination instruments. Applying a systemic
vision, we seek the correlation between the "informatising" system P (that produces
the RbMb support infrastructure) and the "informatised" system B, to facilitate the
global management of the "informatisation" phenomenon: BRbMbP. The crisis of
sporadic engineering can thus be surpassed, shifting from the "reengineering" and
"agile" paradigms to a continuous engineering, where the target R(t) evolves within
the B(t)Mb(t)Rb(t)Mp(t)P(t) system, where production, organization and instruction-
form a global physiology. But such organisation formulas have a cost… that also
must be managed.

Keywords: software engineering, requirements engineering, use case activation,
instruction-production workflows ("functions"), knowledge (carriers) management,
learning by co-doing, systemic and evolving approach, managing the "informatisation".


1. Introduction

In the past years, we have been studying the support of productive-instructive
processes, in a sequence of LICEF projects (ADISA, GEFO, LORNET etc) based on the
development of exploratory prototypes. As we progressed in the definition and
implementation of behavioural specifications for procedural models, active and indexed
semantically, we realized that the instrument produced during this long research-
development process could have also sustained the management ... of its own evolution.
This recursive situation first led us to the idea of applying the "functions" mechanism to
activate the "use cases" employed in SE, then to model and orchestrate application
University of Bucharest and Ovidius University of Constanta


186
development processes through functions and finally to correlate the "informatised"
process and the "informatising" process, by interlacing the use of functions and
metafunctions. Before explaining the systemic method for managing workflows with
instructive potential that we propose here for the software industry (chapter 3), we will
succinctly relate (chapter 2) the vision on which the "function" procedural aggregation
formula is based (Rosca and Rosca, 2008; Rosca, 2006a; Rosca and Rosca, 2004).


2. Facilitating processes- productive or instructive, emergent or orchestrated

2.1. To do and to learn

Human action has exterior and internal causes and effects. Doing something
requires the application of knowledge, and enriches competencies. Learning by practicing
means to do in order to know better, and that- for doing better. An indissoluble spiral is
created between learning and operating - on which the evolution of everybody's
performance ascends. The intimate interlacing between production and learning is found
in communities as well, both at the level of their parts (human participants bearing
evolving meaning, documentary resources loaded with messages, instructively organized
actions) and at that of the unitary metabolism of the respective systems, that do/learn,
produce/evolve. The production and use of resources is indissolubly interweaved with the
production and use of information/knowledge, forming generative cascades: objects and
knowledge used for the generation of objects and knowledge used for ...(Rosca, 2006a).
Interlacing knowledge evolution and activity physiology poses a challenging
theoretical and practical problem: how should we correlate methods and instruments for
action facilitation with those for learning facilitation, surpassing the traditional
separation, through the elaboration of mixed and flexible systems for formative action
facilitation? The coordination logic of cooperative activities (like CSCW), even
specialised for pedagogical cooperation (in CSCL), generally divides the operations
(what the teacher does, what the student does) and is less dedicated to the instructive
sharing of the same operation. Yet, co-action execution in expert-novice pair is the most
natural way to interweave two production/knowledge spirals, so that the expertise of the
one that does because he knows support the one that discovers because he does (Rosca I,
1999). On the other hand, many instruments pretending to be "CSCL", by lacking a
"semantic layer" (not managing explicitly the knowledge involved in the operational
stream) - do not allow learning to be monitored and assisted with the help of computer
networks. They are dedicated to the operational management of a "pedagogic" workflow,
not to the pedagogic management of an operational workflow (Rosca and Rosca, 2004).
Advancement in a cooperative-instructive workflow can be governed by three interlaced
logics: the productive logic of correct operation sequencing, the administrative logic of
intervention coordination (according to rights, responsibilities, availability, contracts) and
the informational (instructive) logic of knowledge application and enrichment (Rosca and
Rosca, 2008).
The 3
rd
International Conference on Virtual Learning, ICVL 2008

187
2.2. System, process, model, lifecycle, reproduction, adaptation, evolution

To usefully influence the work and evolution of a system (material, biological,
anatomic, social, economic, conceptual), one must keep track of its global coherence and
treat it as a unitary whole, with its morphology (structures) and physiology (processes)
determined by its components and the relationships between them (Zadeh, 1969; Rosca,
1999). For mixed systems, in which objects, humans and concepts intervene, the involved
aspects' complexity make tracing, controlling, and optimizing difficult (Le Moigne, 1987;
Andreewsky, 1991). We make use of modelling structures and processes, especially when
their reproduction is sought. Phenomenal reality inspires the model which in its turn
inspires the production of replicas (secondary, derivate phenomena). We can thus create
structures or generate process – similar to the original. Like other resources, structure or
process models have a "lifecycle" (edition/conception, adaptation to various contexts,
usage by those who produce structures or processes on their grounds, annotation for
resuming the cycle) – which constitutes a meta-process, that at its turn can be modelled,
assisted and reproduced.
Analysis and design difficulties rise when the morphology and physiology of the
system, modelled in order to be understood, assisted or reproduced, changes in time,
through the modification of material and cognitive components, or that of the
relationships between them. The system that evolves, adapts, learns- has a plastic
structure, a "longitudinal" existence (Lehmann et al, 2002). Furthermore, it must be
considered in the context of the evolving metasystems it is part of, taking into account
exterior influences and interactions with other systems. One such growth is support
intervention: adding production and cooperation instruments (equipping), reorganising
operations, human assistance, documentary resource supply, increasing participant
expertise through instruction, etc. By enriching an assisted system B with a support
system Sb, we create a new system, BSb – the physiology of which, containing
improvements, can differ from the initial one, more or less.


2.3. Emergent and orchestrated modes: functions

Long term evolutions (Mens et al, 2005; Lehmann et al, 2002) and operational
chains (dedicated to instruction or with productive goal and implicit instructive
dimension) can take place emergently. Participants establish what to do next at each step
(in order to meet requirements, respect conditions and criteria, solve problems) and
choose their support tools (using repositories of available material and human resource).
In order to facilitate the retrieval of informative/instructive resources, the records of
catalogued persons and documents can be indexed: competencies related to some
knowledge, pedagogical potential, communicational particularities, pragmatic considerations.
In other cases, participants to a productive/instructive process conform themselves,
more or less accurately, to pre-established scenarios, that suggest (impose, evaluate,
facilitate) the accomplishment and sequencing of operations. To equip this second,
"orchestrated", mode, combining coordinative, administrative and instructive logics, we
have developed, in the GEFO project (Rosca and Rosca, 2004), the prototype of the
University of Bucharest and Ovidius University of Constanta


188
"function" manager (a biological metaphor, suggesting the physiology of knowledge
management, within intelligent collective organisms).
As can be seen in figure 1, the first step of a "function's" lifecycle is to model the
observed or imagined process. A generic model is defined, composed of abstract
elements: "operations" (represented with ovals, that will become executions), "actors"
(represented with hexagons – that will be concretised with participants) and
"instruments" (represented with rectangles – that will be concretised with resources). The
choice of real elements (using specialized repositories) can be organized in a distinct
phase (adaptation to a beneficiary's context) – or can be distributed in time, from edition
to the execution (enactement, see Ventroys and Peter, 2003) which concretises the
operation. The prepared facilities are exploited, by manipulating, in parallel, the R
resources and the functional model- while resolving ergonomic issues (through
presentation of both windows, or "hiding" resources behind the steering model, or hiding
the supervising model behind the resource's window) and technical interoperation issues
(between the interface, business and data layers of the R and M applications). Based on
execution results (traces, productions and annotations), reports can be created,
evaluations made and corrections directed.



Figure 1. Functions and metafunctions

The function processing chain, represented in figure 1, being at its turn a process,
can be handled functionally, thus obtaining "metafunctions", useful in the global
management of procedures Pb used for supporting the procedure B. Complex
(instructive) functions or "operations" (interfaces for realising a unique activity) can be
The 3
rd
International Conference on Virtual Learning, ICVL 2008

189
managed as any resource, having characterisation records- including competency
descriptions (what someone must know to approach them and what he learns by
executing them). Offering fine-grained cognitive interconnection services requires a
proper preparation (internal indexing) of functions (Rosca, 2006; Rosca, 2008). The
intervention of a semantic service (connection, verification) agent for any sub-operation
of a function (workflow of a modelled procedure) requires some preparing steps: 1.
Indexation, related to the involved knowledge K, of the action to be realized o(k):
required level- coi(k) and level obtained by practicing- cof(k) 2. Presumed competence for
the executing (beneficiary) actor- b(k). 3. Pertinence of the presumed support documents
d(k): from what level cdi(k) to what level cf(k) they can guide the user. 4. Pedagogical
competency of the a(k) presumed assistants, indexed by the (cai(k), caf(k)) pair. The
competence equilibriums monitored are related to the relationships between these abstract
elements and the components that progressively concretise them. For instance, when
concretising an abstract beneficiary b with a person P, having a cp(k) competency, we can
verify if the competency condition required for realizing the operation is satisfied,
depending on the other components that have already been chosen (assistants, documents)
and on the presumed competencies of the still unconnected components. Thus can be
sustained services such as: signalling competency disequilibria, matching a participant
that could optimise assistance etc.


3. Application: the systemic engineering of "informatisation" workflows

3.1. Progressive use-case activation in software engineering

Modelling processes that take or should take place within a technically assisted
social system (Herrmann, 2004), requiring representation and coordination techniques for
man-machine orchestration, was intensely approached in software engineering (Hermann
at al, 2004; Noelle et al, 2002). Between the founders of UML, Jacobson was the one
who promoted standardising the modelling of sought processes ("use cases") as central
design instrument (Hollub,2001; http://www.uml.org). In the RUP methodology (Larman
et al, 2002; http://en.wikipedia.org/wiki/IBM_Rational_Unified_Process), an application
is developed by continuous reference to these physiological descriptions (see Fowlers'
observations at http://www.martinfowler.com/articles/newMethodology.html#N401).
Apart the dialogue between architects and developers, the relationship with a client
imposes the description of the aimed processes, in textual or graphic languages, more
natural-culturally, exploited in "scenario-based design" (Bustard et al, 2000; Caroll,
2000). In "requirements engineering" specification extraction techniques have been
elaborated, including use case management methods (Anton et al, 2000).

University of Bucharest and Ovidius University of Constanta


190
In this context, model activation formulas, with the help of the "function manager",
can intervene. The proposed method is described (illustrated) in figure 2:

R
B scenario
3 Rb
activity B
equipped
with Rb
facilitated
with Mb
2
4, 8
activate
admin/support/ instruction Mbrsai
Rp
R
(Cybershop)
r
use case Mb
man
Cybershop R
add
testbed Mbr
de
3
prototype
6
develop
9
refine
ad
5 test
re
2
compose
se
architectur e
+dev plan
7
deploy
1
activity P
1 edit
B test process
R
catalog
user admin
cons shop
order
tr eat
shcard
man
Cyber shop R
add
catalog
user admin
cons shop
order
treat
shcar d
asis1
asis2
supp1
adnot1


Figure 2. Lifecycle of an active use case

Let us consider a system B – that will be "informatised" by introducing an Rb
resource (program, application, infrastructure) – transforming the system into BRb. The
Mb modelling of the current, un-equipped processes (or Mbr of the aimed processes,
equipped with Rb) is usually done in passive forms. Three distinct phases are sequenced:
modelling (the B-Mb relationship), development (the Mb-Rb relationship) and use (the B-
Rb relationship). By enriching the procedural model Mb with facilitation layers for
participation to B processes (including access intermediation to the built resource Rb), we
can go from a simple "use case" to an orchestration function –- Mbro, a facilitation
interface, placed between B and Rb. Such use-case processing permits a coherent
management – in RUP vision – of the entire lifecycle of a software product, from the Mb
edition phase to the Rb production phase (based on Mb) and afterwards to the use of an
activated model Mbr as an interface for prototype testing – Mbrt, or as a tool for user
support- Mbrs, operation administration – Mbra or user instruction – Mbri.
Observing (or imagining) a B application scenario (for instance – the use of a
virtual store by those responsible of updating the product catalogue, then by the end-user
that adds products to the shopping cart and places an order and finally by the
administrator that handles commands) – we can edit (phase 1) a first form of an Mb "use
The 3
rd
International Conference on Virtual Learning, ICVL 2008

191
case", a passive model of the B operational chain, that can be used as inspiration source
by the conceiver of the desired application's architecture (phase 2) or by the prototype
developers (phase 3). Afterwards (phase 4), by adding functional facilities (sequencing,
coordination, annotation, binding to existing resources and to the tested prototype), we
attain a facilitating or orchestrating use case – Mbr or Mbro – upon which can be
organized the activity (phase 5) of testing the procedure assisted by the evolving
prototype Rb. The development cycle (phase 6) – (or the modelling) can be resumed, for
the progressive improvement of Rb – fuelled by the data captured by the Mbrt test device.
After finalizing Rb's development and installation in the beneficiary's context
(phase 7), we can operate (phase 8) a new functional activation (preparation) of Mbr,
leading it to an Mbra management instrument (model centred on the enrichment of
administrative logic- access and surveillance) or to an Mbrs support instrument (centred
on the assistance facilities). The Mbr use case becomes a facilitation tool for the Rb
application's use within the beneficiary system B. Preparing instructive co-action
mechanisms and semantic services, the final activation can transform Mbrs from a
support instrument to an instruction instrument Mbri (of tutorial type, with human
assistants and support documents or with automated advises). An Mbrasi model – rich in
all facilitation layers – will provide coordination, administrative, support and instructive
services – and thus will be usable as a plastic form of support for the B system,
optimising the added infrastructure's use. Tracing and annotation capabilities
intermediated by Mbr can also be used for the inspiration of refinements and corrections
to the RbMb application (phase 9).


3.2. Active modelling of development workflows

Building resource-applications Rb, passive models Mb or active interfaces Mbro is
accomplished in a software production process P, that exploits specific resources. Figure 2,
for example, illustrates a programming method P based on the "functional" management
of use case evolution. However, we can also resort (figure 3) to a Mp functional
modelling of the proposed method's flow, usable for the orchestration (management,
coordination) of P programming activities or for orienting the conception of an Rp
resource, necessary to programming. Workflow management is now both target (in
system B) and instrument (in system P).
Software engineering is dedicated to optimising processes that take place in the P
system. Facilitating these processes can be accomplished through: equipping,
reorganization, assistance, instruction. We can again compose Mp models and construct
Rp applications that can be accessed through activated versions of models – Mpr, having
enriched coordination, administration or instruction logic layers. Making abstraction of
the beneficiary systems B-n (we'll come back on this in 3.3), let us observe that activating
the programming workflows Mp implies new functional cascades: reality-model-
construction-application. The considerations presented in 3.1 can thus generate, through
particularisation (specialisation), specifications for programming support instruments,
based on the activation of models representing design procedures. We reach the
management of a "programming case's" lifecycle:
University of Bucharest and Ovidius University of Constanta


192

potential activities
Bn assisted with
Mbn
P activity
managed
with Mp
de
r e
se
ad
archit
+plan
r
scenarios Bbn*
Rb
ut
ur
tests Bn*
r
Rbt
r
use case Mb
testbed Mbr
r
10
edit Mp
RE
14 refine and apply
generic Mp
DE
4, 8
activate
Support
Mbs
r
Rp
r
use case
Mb
testbed
Mbt
de
3
prototype
6
develop
9
refine
5 test
r e
2
compose
se
ar chitecture
+plan
7
deploy
r
Mpits
ad
1 edit
SE
AD
12
execute
Mp
Rp
requirem
specify
archit
elaborate
Res
11activate
Mpr
AD
develop Rp
u
r
13 testbed Rp,
calibr ate Mpr


Figure 3. The lifecycle of an active programming model

The Mp software production process's model is obtained through manual edition
(phase 10), but its extraction by interaction interception – within a pilot (demonstrative)
programming process – is not excluded. The sequencing and coordination logic of the
(team) programming process – expressed by the Mp functional workflow – can
correspond to methods of work organization in SE, be subject to standardization (norms)
for production quality evaluation or facilitate the project continuation by other persons or
teams etc. By an Mpr correlation between Mp and Rp and the activation of the model
(Mpro), we can coordinate the programming activity and facilitate the use of the Rp
instruments, eventually rebuilt upon the Mpr model (phase 11). The execution of these
development scenarios (phase 12), for building various Bn applications – exploits the
facilities organized during the preparation phase of the Mp development model.
Enriching the Mp model can also be accomplished, in order to test the Rp
instruments or calibrate the P method by creating (phase 13) Mprt models, allowing the P
development method to be tested. These calibration tests, supervised by an AD meta-
administrator, can orientate the modification of the programming workflow or of an Rp
programming instrument. The development of Rp, conceived by the DE meta-developers,
can be resumed, upon the meta-architecture elaborated by SE meta-engineers, based on
the behavioural specifications defined by the conceptual researchers. After refining the
development method and the design support instruments, we can implement (phase 14)
The 3
rd
International Conference on Virtual Learning, ICVL 2008

193
new programming management facilities, by processing (activating) Mp models so that
they become Mpr tools, specialized as "test", "support" or "tutorial".
Before applying generic organization functions to the development of a particular
application B (see 3.3), we recall that the flexibility of the Mpr model permits assistance
metamorphosis between: action inspiration and guiding, project advancement observation,
performance analysis and evaluation, information and learning, communication and
cooperation coordination, adapted human and documentary support, locating, accessing and
manipulating programming instruments, operation automation, intervention management –
according to rights, contracts and policies.
All these facilities, but especially the evolving explicitation of cooperative
programming procedures, the operation sharing (so that they can be effectuated in expert-
novice instructive tandem) and the competence explicitation (to keep track of
programming knowledge evolution) – create an appropriate context for an "as you go"
integration method of interns, in programming teams, allowing a smooth transition from
the initiation through instructive participation to the productive participation, with
continuing knowledge development. By refining access logic, we can support the
formation and work of ad-hoc constituted programming teams- according to availabilities
and the existing conventions, in a distributed programmer community (Rosca and Rosca,
2002; Lee, 2003). The optimisation services for the concretisation of human and
document resources bound to the development workflow require the indexation of
person, resource and operation repositories- using an adequate knowledge reference
system. Using the participants' competencies (programming expertise), the support
documents' pedagogical pertinence or the opportunity of available training operations-
competency equations can be solved, resource selections and role allocations made and
other services of expertise management offered.
There are frequent cases (Larman, 2002) when programming activities are too
complex, variable, un-reproducible – to be meticulously planned, efficiently. In such
cases, operation sequencing is done emergently, participants deciding what they do at
each stage, choosing the required resources and respecting a given set of rules. Using Mp
workflows can intervene even in such cases, after realizing productive cascades (to
illustrate them a posteriori), using various observation methods (action interception,
annotations, etc). Organizing knowledge reference systems specific to the programming
activity and indexing resources and expertises upon them – can therefore be useful,
regardless of the production mode (emergent or orchestrated).


3.3. The systemic engineering of "informatisation" workflows

Isolating the programming activity from the physiology of the systems for which it
builds support tools- can be useful, only when we seek optimisation of some generic
processes and instruments. However, the P-n activities of developing Rb-n tools
(including passive Mb-n and active Mbr-n models) for the support of processes within the
B-n beneficiary systems generally depend strongly on the B-n system's metabolism, as
suggested by the Pb indexing. Understanding the functioning of the system of systems:
BMbrRbPb has major consequences for the success of "informatisation" projects.
University of Bucharest and Ovidius University of Constanta


194
Isolating the informatising process P from the informatised process B (reducing the
relationship between them to the – "deliverable" – product Rb that P offers to B) has led
to superficial treatment of the beneficiary's requirements (Rosca, 2006b).
If the beneficiary B, confronted with new situations (changes in objectives,
priorities, equipment, criteria, etc) requires a modification of the Rb infrastructure, a
reconstruction process is run (in which Rb is enriched or modified.). The scale of
"reengineering" needs (http://www.sei.cmu.edu/reengineering/) has caught the software
industry off guard (Larman, 2002), diminishing reproducibility. Hasty approaches lead to
endless requests for corrections. Careful ones delay application to the point where initial
specifications no longer correspond. Teams called on to resume a project encounter major
difficulties in understanding and continuing what has been done in the antecedent
construction, asking for a solid documentation, costly for the initial designer. Supporting
very labile informational systems must surpass the paradigm of cascading requirements/
development/ application) cycles. Socio-economical systems change continuously,
become. An infrastructure that pretends to sustain their physiology should evolve
continuously, ameliorate progressively, become – together with the beneficiary system.
"Continuous software engineering" requires a cybernetic view of the relationship between
the producing system P and the beneficiary system B – forming an evolving meta-system.
For the support application MbRb to evolve together with the B system that solicits and
uses it - you must manage the evolution of Mb and Rb following their coupled long-term
lifecycles. All these can lead to a global management formula, based on functions.

r
scenario B(t)
ub
r(t)
activity B( t)
equipped
with r(t)
facilitated/
managed/
with Mb(t)
ubt
admin/suppor t/ instruction
Mbs(t)
r
use case Mb(t)
man
Cybershop R
add
testbed Mbt(t)
ubr
test B(t)
r
catalog
us adm
cons shop
order
treat
shcard
res
Cybershop R
add
catalog
us adm
cons shop
order
tr eat
shcard
as1
as2
sup1
adno
4, 8*
activate
Mbrsai
r
Rp
r
use case
Mb
testbed
Mbr
de
3*
prototype
6*
develop
9*
refine
ad
5 *test
re
2*
compose
se
architecture
+dev plan
7*
deploy
activity P
managed
with Mp(t)
1* edit
de
re
se
archit
+plan
Rp
r
Cybershop r(t)
ubr
as2
ubt
ad


Figure 4. Global management of the informatisation process's evolution

The 3
rd
International Conference on Virtual Learning, ICVL 2008

195
Figure 4 illustrates the procedural flow of the global evolving system
BMbrRbPMprRp(t) – presuming that the represented processes are resumed, as many
times as necessary (see the * symbol). The beneficiary system B has an evolving
morphology and physiology B(t), even the continuous improvement of the Rb(t)
resources it employs – being a transformation. We can establish multiple connections
between the components of the B and P systems (managing the interference between B
and P processes- being the goal of model 4). It could be the communication between two
persons from the B and P spaces, their coordination in common processes, resource
sharing, mutual support – with the occasion of collaboration in the Mb(t) scenario's
definition, solution testing with the help of the Mbrt(t) model, Rb(t) product installation
or the analysis of data collected by the support interface Mbrs(t).
Processes in B are observed in a convenient rhythm, and used by specification
engineers to edit and modify the Mb(t) models – following observations made in the
method's upstream (but downstream in the temporal evolution spiral sense). Using these
models, developers, organized within the P(t) evolving physiology programming process –
program or reprogram the Rb(t) instruments. Also within the P system takes place, in
several rounds, the Mbr(t) test models' activation (reactivation), by enrichment with
improvement layers of the coordination, negotiation and annotation logic. The active
Mbr(t) models, installed at the beneficiary, along with Rb(t), and used for the
inspiration/orchestration of activities in B, facilitation of work with Rb instruments,
participant connection or instruction – can be progressively improved. Interactions
between the Bx components of the B phenomena and the Py components of the P
phenomena can be realized by sharing the Rb product, through the Mb or Mp models or
outside them: BxPy, BxRbPy, BxMbPy, BxMpPy, BxMbRbPy, BxRbMpPy, BxMbMpPy, etc.
The P process cascade that supports the B(t) system's evolution by equipping it
with an evolving Rb(t) support can be modelled in passive forms Mp or orchestrated with
active Mpro models – that may facilitate the integration of new persons in the
development team or the transformation of the Rp programming instruments. As the
software industry is in continuous "progress", the activity organization formula within the
P system is also evolving (see chapter 4.2). Another cause for P's organization change
can come from the requirements emerging from B – that can't be optimally satisfied
anymore by the old development formulas. Processes represented in the P layer must
therefore evolve, adapt, sustain a labile P(t) physiology, axed on the evolution of Mp(t)
models and on the use of programming instruments in continuous renewal: Rp(t).
Figure 4 thus represents an evolving engineering, based on the continuous update
of Mb(t) models and Rb(t) resources – within evolving design procedures Mp(t). The
global bi-layer stream represents the correlated history of the informatised –
informatising system BMbRbPMpRp(t). By organizing it as a function (enriching the
coordination, negotiation, assistance and instruction layers), we can obtain an
orchestration tool for the informatised-informatising ensemble, a management instrument
for P design workflows that supports some B use workflows. Modelling such a
management mode could at its turn be represented or piloted by higher-rank models. This
"organic", recursive and "genetic" approach leads us, in the SISIF project, to a continuous
engineering of informatisation: the modification of MbRb(t) in the labile Pb(t) workflow
so that B(t)'s optimal support can be attained, along its evolution.


University of Bucharest and Ovidius University of Constanta


196
4. In guise of conclusion: the problem of costs

Informatisation is not limited to the development of applications. It also consists in
the amelioration of the informational physiology, the informatising process continuously
supporting the evolving system in which the informatised processes take place. A process
can be assisted by facilitating the access of participants to resource repositories or by
planning passive or active workflows. For the computer network to operate "synaptically"
(Lee, 2003) matching partners (and then eventually supporting their cooperation with
workflows having pedagogical potential) – a "knowledge" explicitation is necessary.
However, the organization of these facilities involves costs and is not profitable if a
certain reproducibility does not occur. Benefits must surpass the preparatory expenses
(Olufunmilayo, 2006). The problem of economic management of knowledge evolution
(Rosca, 2008a) is complementary to that of the economy of operational workflow
management (Rosca, 2008b) Informatisation (automation) projects should not be
accepted without doubts, from technological enthusiasm or blind obedience towards the
obligation of updating- pushed by an imaginary competitiveness (Rosca, 2006b). Huge
funds have already been expended for "necessary" transformations with unknown local
and global value, neglecting the real profitability issue, or lacking an appropriate
analytical apparatus.
The stratified question we conclude our analysis with is:

on what analytical grounds can we solve economical problems such as:
"Is it worth (according to chosen criteria) to organise resource repositories, model
and activate workflows and explicit knowledge at a certain granularity level – in order to
support a class of productive-instructive processes/applications, that will take place in a
system, evolving in a certain period?".

We have been confronted to this issue on the course of our participation to the
LICEF chain of projects (that have raised many questions regarding the efficiency of
instruction technologisation efforts)- decomposing it in a few questions: "How profitable
is the meticulous management of metadata records for components that can participate to
the processes of a socio-technical system?"; "To what degree is the modelling of
processes that take place in the system, or the activation of these models (to be used as
orchestration instruments) – profitable?"; "Is the explicit management of knowledge and
competencies advantageous in comparison to the case where their evolution intrinsically
results from the activity of knowledge bearers?".
We have sought a synthetic expression for these questions, in the form of
"profitability equations".

1. Equipping emergent operations. In comparison to unprepared emergent
(programming) operations, preparing Rp material and Hp human resources (metadata
record catalogue, etc) introduces additional management costs Cmr and Cmh. The
profitability condition in this case is that the additional gain G be greater than the
organization expenses:

The 3
rd
International Conference on Virtual Learning, ICVL 2008

197
(1) Cmr + Cmh < G ?

2. Semantically supported emergence. When resorting to the explicitation of
persons' competencies Hk and pertinence of documentary resources Rk relating to K
knowledge (reference prepared by the edition of knowledge structures and competency
norms used as reference system), additional expenses are introduced for the semantic
support of emergent operations: managing knowledge Cmk, human competencies Chk
and documentary pertinence Crk. Facilitating resource retrieval – brings a Gk benefit.
The profitability condition is that the additional benefit (in contrast to using semantically
unindexed repositories) surpass the additional expenses:

(2) Cmk+Crk+Chk<Gk-G?

3. Operation orchestration. Another possible improvement, is the use of activated
operation models ("functions") O, that facilitate the inspiration of sequencing, access
management and resource manipulation, collaboration coordination, etc. In this case, the
profitability condition is that the additional gain (to the simple emergent case) Go-G
surpass the Cmo expenses of organising operational models (edition, activation, etc):

(3) Cmo<Go-G?

4. Semantically supported orchestrations. The last situation is that where both
repositories and operational models O are semantically indexed on K reference systems.
The realized benefits (possibility to offer optimization services, matching services, etc)
Gok must justify the Cok expenses for procedural model indexation:

(4) Cok<Gok-Go?

In the absence of a "magic formula" for solving these open problems, we can
formulate some orientations:
1. Estimating the global profitability of a global engineering approach requires a
rigorous meta-analysis (using intuition, trends and slogans being legitimate… only when
the cost of meta-analysis is – too – prohibitive). 2. The analysis must begin with the clear
enunciation of the priorities (values) chosen by those involved in defining the
specifications. 3. The methods can be judged and combined, according to the particular
context: Are we dealing with a large scale application – posing retrieval problems – or
with a small scale one – allowing easy participant orientation? Is the need for
discretion/confidentiality a priority, or that for tracing/evaluating the activities and
knowledge evolutions? Do we wish to conserve the system and only ease some
operations or to ameliorate the physiology? Is the involved knowledge stable, or in
continuous change? Do collective, or individual interests, take precedence? Spiritual, or
material goals? Minimum price, or maximum quality? Do we want to exploit human
intelligence or automatisms?
When the reproducibility of the assisted operations is substantial (see situations like
Google, EBay, Amazon, etc), efforts for preparing retrieval and matching, based on
semantic inferences – are justified. For unrepeatable phenomena, with rapidly changing
topology and physiology – the investment in semantic explicitation (to permit computer
University of Bucharest and Ovidius University of Constanta


198
intervention) will not redeem, exploit the participants' intelligence being more profitable.
In intermediary situations, the problem may become undecidable, due to barriers created
by complexity and even by epistemological difficulties.
The problem of managing… management costs (also see comments at http://www.strassmann.
com/pubs/measuring-kc/) remains provocative. We will try to deal with it as "recourse to
the method": modelling the estimation process.



REFERENCES


ANDREEWSKY, E. (1991), Systemique & Cognition, Dunod, Paris, 1991.
ANTON, A., DEMPSTER, J., SIEGE, D. (2000), Managing Use Cases During Goal-Driven Requirements
Engineering, http://www4.ncsu.edu/~aianton/pubs/icse2000.pdf
BUSTARD, D. W., HE, Z., WILKIE F. G., (2000), Linking soft systems and use case modelling through
scenarios, Interacting with computers 13 97-110, 2000, Elsevier Science.
CAROLL, J. M. (2000), Five reasons for scenario based design, Interacting with computers, 13 43-60, 2000,
Elsevier Science.
HERRMANN, T., HOFFMANN, M., KUNAU, G., LOSER, KAI-UWE (2004), A modelling method for the
development of groupware applications as socio-technical systems Behaviour & Information
Technology Vol. 23, No. 2, 119-135, 2004, Taylor & Francis Ed.
LEE, Y, CHONG, Q. (2003), Multi-agent systems support for Communities-Based Learning, Interacting with
computers 15, 33-55, 2003, Elsevier Science.
LARMAN, C., KRUCHTEN, P., BITTNER, K. (2002), How to Fail with RUP Process: Seven Steps to Pain
and Suffering, http://www.agilealliance.org/articles/larmancraigkruchtenph/fi
LEHMANN, M. M., G. KAHEN & J. F. RAMIL (2002), Behavioural modelling of long-lived evolution
process- some issues and an example. Journal of software maintenance and evolution: research and
practice, 335-351, 2002, John Wiley & Sons Ed.
MENS, T., WERMELINGER, M., DUCASSE, S., DEMEYER, S., HIRSCHFELD, R., JAZAYERI, M.
(2005), Challenges in software evolution, Principles of Software Evolution, Eighth International
Workshop on Volume, Issue, 5-6 Sept. 2005, pp. 13-22.
LE MOIGNE, J. L. (1990), La modelisation des systemes complexes, Dunod, Paris.
NOELLE, T., KABEL, D., LUCZACK, H. (2002), Requirements for software support in concurrent
engineering teams, Behaviour and information technology, Vol. 21, No. 5, pp. 345-350.
OLUFUNMILAYO, B. A. (2006), "Measuring and Representing the Knowledge Economy: Economic Reality
under the Intangibles Paradigm" Buffalo Law Review 54 , 1-102.
ROSCA, I. (2006a), Knowledge and object phylogenetic production cascades – the TELOS case, CE'2006
conference, Antibes, Proceedings in Leading the web in concurrent engineering, Parisa Godus & Others
Ed., 2006, pp. 296-304, IOS pres, 2006.
ROSCA, I. (2006b), The risks of virtual learning ICVL'2006 Proceedings, Bucharest Univ Press, Marin Vlada
& Others Ed.., pp. 155-163.
ROSCA, I. (1999), Towards a systemic vision of the explanation process; the story of a research on integrating
pedagogy, engineering and modeling- PhD thesis, Montreal.
ROSCA, I. and ROSCA, V. (2008), Meaningful access: policy, management and orchestration. IJAMC
(in print).
ROSCA, I. and ROSCA, V. (2004), Pedagogical workflow management with functions, LOR'04
congress, Montreal.
ROSCA, V. (2008a), Explicit knowledge management and management of knowledge carriers.
Graduation Paper -Economy-Management, Iaşi, 2008.
ROSCA, V. (2008b), Connections between expertises, activities and documents in Knowledge
Management, Graduation Paper -Management, Iaşi, 2008.
VANTROYS, T., PETER, Y. (2003), Cow, a flexible platform for the enactment of learning
scenarios, CRIWG proceedings, Autrans, France, 2003, Springer Verlag.
ZADEH, L. A. (1969), System theory, McGraw-Hill, N.Y., Toronto.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

199
Steps of Implementing an E-learning
Programme in Superior Education

Gabriela Moise
1
, Loredana Netedu
1
, Liviu IoniŃă
1


(1) Petroleum-Gas University of Ploiesti,
no. 39 Blvd. Bucuresti, Ploiesti, ROMANIA
E-mail: gmoise@upg-ploiesti.ro


Abstract
This paper presents a new definition of e-learning, as a blend of the traditional
classroom-based learning with the internet-based one. It takes into consideration
the fact that learning is more comprehensible than training, as it includes
unintended and non-institutionalised learning as well. The authors consider the
Internet technologies a means suitable for creating and delivering an instructional
environment and not a purpose in itself. Taking into account the above-mentioned
opinions, there are defined and described the steps to be used in implementing an e-
learning programme in superior education, steps such as: establishing the target
audience, designing the traditional lecture and its electronic version, as well as their
curricula, establishing a relative, adaptive percentage of the two components
involved, designing or/and accessing virtual communities related to the approached
field, purchasing the necessary, most appropriate technology. An application, now
in progress at the Romanian-English section, within Petroleum-Gas University of
Ploiesti, is presented in the final part of the paper.

Keywords: E-learning programme, blended learning.


1. E-learning: An Ever-changing Concept

The term E-learning suffers a lot of definitions. It comprises online learning, virtual
learning, web-based learning, and so forth. Nowadays, any computer is connected to a
network, fact that implies using the term E-learning in a broad sense.
Therefore, we select some definitions of the E-learning concept.
“E-learning is the use of the Internet technologies to create and deliver a rich
learning environment that includes a broad array of instruction and information resources
and solutions, the goal of which is to enhance individual and organisational
performance.” (Rosenberg, 2006)
“E-learning would incorporate all educational activities that are carried out by
individuals or groups working online or offline, and synchronously or asynchronously via
networked or standalone computers and other electronic devices” (Naidu, 2006)
“We define e-learning as instruction delivered on a computer by way of CD-ROM,
Internet, or intranet with the following features: includes content relevant to the learning
objective, uses instructional methods such as examples and practice to help learning, uses
University of Bucharest and Ovidius University of Constanta


200
media elements such as words and pictures to deliver the content and methods, builds
new knowledge and skills linked to individual learning goals or to improved
organisational performance.” (Clark and Mayer, 2006)
The evolution of the information technologies has generated changes in the
meaning of the term E-learning, and the directions of the computer science development
have drawn new directions in the E-learning development. Experts of the instructional
design, teachers, universities leaders and students are permanently challenged to adapt,
reconfigure, redesign the traditional education so that it may cope with to the “e” era or,
more recently, to the “m” (“m” for “mobile”) era. All the above-mentioned definitions
state that there is an electronic medium of learning delivery, two definitions (namely, the
first and the third one) focus on increasing the organisational performance, whereas the
last definition also includes the instructional techniques and pedagogical content. In our
opinion, E-learning consists in any type and form of learning delivered via electronic
devices, be it intended or unintended, online or offline, synchronous or asynchronous,
formal or informal learning. It can be better applied to adult learning, and, by combining
face-to-face learning with E-learning, one may obtain what we call blended learning.
A sound learning is the one that satisfies the needs of heterogeneous groups of
people, and, therefore, blended learning implies a mixture of different learning
approaches in the attempt to obtain the best outcomes possible. We apply these strategies
at our courses so that we may bring the liberty of expression from the virtual environment
into the classroom. In the paper entitled Strategies for building blended learning (Rossett
at al, 2003) the authors state that blending involves a planned combination of approaches,
as follows: a supervisor’s coaching, participation in an online class, breakfast with
colleagues, reference to a manual, online communities.
The criteria for determining the optimal degree of learning support are: learner’s
characteristics, learning’s outcomes, learning’s activities, context of learning, hardware
and software infrastructure, time management, teachers’ experience.
This paper will further present in detail the steps of implementing an E-learning
programme used in our teaching activity.


2. Implementation Models Now in Use

The design of the online instructional system has been realised by a number of
professionals in the instructional design of computer-assisted instruction (Kemp, 1994;
Clark, 1995; Dick & Carey, 1996). There are a lot of models of ISD (Instructional System
Design), but most of them are based on the ADDIE model (http://ed.isu.edu/addie/).
ADDIE model consists of five phases, each of them having outcomes which become
inputs for the next phase.
Each phase of the ADDIE model implies a series of steps that lead to the model’s
application in the online instructional process. Briefly describing, the ADDIE model
consists in: the analysis phase, within which the goals and objectives of the instruction
are established, the learner profiles are identified, as well as the available technologies;
the design phase, when the instructional process (strategies, techniques) is designed, the
The 3
rd
International Conference on Virtual Learning, ICVL 2008

201
development phase, in which the electronic courses are built, the implementation phase,
when the instructional system is delivered, the facilitators, the trainers and the learners are
trained, the evaluation phase, during which the quality of the product is determined and
the collected feedback generates ideas to improve the system. MRK (Morrison, Ross and
Kemp, 1996) model consists of nine steps:
1. identifying the instructional problems, and specifying the goals for designing
an instructional program,
2. examining the profile of the learner that should be paid attention to during planning,
3. identifying the theme of the content, and analyzing task components related to
the stated goals and purposes,
4. defining the instructional objectives for the learner,
5. sequencing the educational content within each instructional unit for a logical learning,
6. designing the instructional strategies so that each learner can reach the objectives,
7. planning the information delivery,
8. developing the evaluation tools to assess the objectives,
9. selecting resources to support learning activities.
The original diagram of Kemp model can be found at the address http://www.personal.
psu.edu/users/s/j/sjm256/portfolio/kbase/IDD/images/kempmodel.jpg.
A comparative study on all these models leads to the following conclusion: most of
ISD models have a systemic pattern. In the initial phases, the objectives are identified and
the profile of the learners is defined. On the minus side of ISD models, one can mention
the impossibility to mould the learning process while in progress.
Generally speaking, the E-learning programme implementation complies with the
following steps: a preliminary analysis, consisting in specifying aims and learning
outcomes, learners’ analysis and context analysis, programme’s development, its testing
and implementation and programme’s evaluation and improvement.


3. Implementing an E-learning Programme in Superior Education

3.1. Definition

The programme that we intend to implement is in fact a blended learning
programme. Here are some reasons for preferring it to of an exclusively traditional face-
to-face course or an exclusively e-learning one: nowadays students get hired during
university and, because of this, they do not come to classes or they do not have time to
study; one may also notice a lack of motivation and void of interest as far as their
education is concerned, but interest in using the Internet. It is a reality that students from
the first years do not know how to learn, so their metacognitive abilities have to be
improved.
We can attract students with more appealing courses, i.e. more flexible and
provocative and less stiff, as the traditional lectures are usually considered.


University of Bucharest and Ovidius University of Constanta


202
3.2. Goal

Our goal is to use it in superior education so that we may increase students’
performances. Teachers are due to reach the objectives of the course in given conditions
and in a specified time. A graphical representation of the most likely evolution of the
students’ knowledge is represented in figure no 2. There is an initial state, that of the
students’ previous knowledge and a target that is to be reached within a specified time,
i.e. 14 weeks. In our opinion, an e-learning programme should increase the chances to
reach the established goals, or, at least, to get the students closer to the target. O
1
, O
2
, O
3

stand for courses’ objectives, and K, for competence, i.e. knowledge and abilities, level.



Figure 2. Graphical Representation of Students’ Knowledge Evolution
The 3
rd
International Conference on Virtual Learning, ICVL 2008

203
3.3. Steps

In their rush to implement E-learning programmes, some educational organizations
make mistakes because of the inadequate calculation of the financial, time or human
resources. Much money and more time have been wasted in the attempt to achieve a
successful E-learning programme. We propose a seven-steps methodology to implement
an E-learning programme.

1. Analysis phase
• Analysing the learning content. There are five types of content in learning:
facts, concepts, process, procedures and principles. The primary types of content are
called artefacts of knowledge. (Clark, Mayer, 2002)
• Analysing the goals. There are two types of goals: to inform and to perform.
The benefits of goals’ analysis and setting are: students are more motivated to learn, they
will know why they have to learn a procedure and where they can apply it, all projects
can be evaluated according to the results obtained by each student. At the end of a
training session one can report the students’ outcomes related to the proposed goals.
Nowadays, there is a lack of reports as regards the students’ performances.
• Analysing human resources. To build an E-learning programme requires
specialists from different areas: teachers, instructional designers, IT experts: web
designers and programmers, software developers, computer networks and Internet
specialists. In most cases a lot of E-learning projects have failed because of the lack of
specialists. It is important to stress the fact that a teacher cannot be an e-course developer.
• Analysing technologies. Don’t spend money on unnecessary technologies! Use
the technologies already existing in your organizations to solve immediate problems and
buy the technology you need in the present and near future! Estimate the software and
hardware technology report on: space area, number of students, number and types of
courses, multimedia content, number of staff, existing headquarters. One needs
technologies to develop and deliver courses and to manage all E-learning programmes.
• Analysing learners context and suitable pedagogical procedures. The context of
the learning process may be: mental, social, technological, emotional, knowledge, and
classroom context. (Moise, 2007) There are a lot of pedagogical procedures that can be
applied in an E-learning programme: learning by doing (scenario-based learning, project-
base learning, case-based learning, problem-based learning, role-play-based learning),
collaborative techniques, blended learning. Our piece of advice is to use the most suitable
procedure(s) for the given instructional objective and learner context.

2. Building the project plan
The success of any project, an E-learning programme included, may be
assured by a well-conceived plan. Use a project management methodology and
use a project management information system! Any project has to be monitored
from planning to operations. Assure a management team responsible for monitoring
the whole project! Make a budget plan, estimate the costs of the E-learning
programme’s development, implementation and maintenance! As for the pedagogical
University of Bucharest and Ovidius University of Constanta


204
strategies, use blended learning, combining the most appropriated procedures to
assure the success of the learning process.

3. Building the prototype
A prototype is an instance of the project built for demonstrating the functionality of
the process, in our case that of the E-learning process. The E-learning prototype is a basic
version of the E-learning system. The prototype is used to assure the success of the E-learning
programme, it validates all the techniques and strategies implied in the programme. Build
a prototype of every e-course! Use a methodology to produce a good e-course and name
an experts’ team to perform this! (Often the techniques to produce a film are used.)

4. Implementing the prototype on a small scale: a group of 20-30 students
Implement more e-courses from different areas and don’t forget to implement the
management system! You have to manage a lot of entities: pedagogical resources,
students, staff, educational plans, grades and registers, fees, financial information and so
forth. In this phase the costs of the implementation (often neglected) may be estimated.

5. Evaluation of the E-learning programme
Evaluate the outcomes of the students and the objectives reached through the
programme! Make a report regarding the incomes and expenditures!

6. The E-learning programme’s correction and extension of the project
Be sure that you allocate enough time for all steps of the project’s development!

7. Assuring the maintenance of the E-learning programme
The core of a successful e-Learning programme is the mode of training and
learning using electronic devices. This paper presents a possible transcription of a
traditional face-to-face lesson into a blended learning one.


4. Experimental Course

We are currently developing and implementing an E-learning programme in
Petroleum-Gas University of Ploiesti. Following the above-mentioned steps, we have
decided to use software and hardware technologies, as well as human resources already
existing in our university in order to transpose a traditional Practical Course activity,
namely Getting to Know You into a blended lesson.


4.1. Traditional Practical Course – English Description

The official description of the Practical Course – English Language 1, an obligatory
discipline included in the educational plan of the Romanian-English Section (functioning
within the Faculty of Letters, Petroleum-Gas University of Ploiesti), establishes four
The 3
rd
International Conference on Virtual Learning, ICVL 2008

205
types of specific competencies that are to be followed throughout university at this
discipline: getting to know and understand information, explaining and interpreting texts,
instrumental and practical competencies and attitudinal ones.
These are further detailed for each year, the competencies for the first year
including assimilating and applying grammar and vocabulary knowledge, expressing
opinions about the selected topics or the ones included in textbooks, receptive and
productive skills development and practice and developing team spirit, responsibility,
indulgence, respect towards others’ opinion. Although one may identify grammar lessons,
reading or writing ones, lessons of revision, a Practical Course activity is usually
designed as a mixed lesson, in which the students are given the opportunity to acquire,
enrich and apply English grammar and vocabulary in real or simulated contexts and in a
permanent interaction with their colleagues and teachers.
The students in this upper intermediate/advanced class are between the ages of
18-22. There are 13 female students and 2 male students. Because one of the two classes
starts at 8 in the morning, students are often quite sleepy, finding difficult to concentrate
and use a foreign language. Otherwise, they are attentive and really interested in
practicing their skills and lexis, let aside grammar they are not very “fond of”.
In the past four lessons the students have been discussing the issues of how one can
use language in his/her advantage for a better communication and overviewing the
indicative mood. They have listened to short extracts in order to match descriptions to
speakers and order some topics. They have been looking at vocabulary and expressions
related to introductions or describing other people, i.e. adjectives and nouns and they
have also been discussing style and register of oral or written accountants. They have
revisited a number of past tenses, including hypothetical past (third) conditionals (‘If he
hadn’t lost his job, he wouldn’t sold his house.’).
Next week the class will start working on a unit entitled “Can You Believe It?”
which includes as structure to be revised tenses in accounts and narratives, writing a
competition entry, error correction exercises, based on the grammar and lexis discussed
up to that moment.


4.2. Aims

To allow students to practise speaking spontaneously and fluently about
something that may provoke the use of words and phrases and grammar they have been
learning recently.
To give students practice in reading both for gist and for details.
To enable students to describe themselves and others, taking into consideration
verbal and/or body-language communication used in real or hypothetical situations.
To have students produce an account by organising information and using a
formal register.


University of Bucharest and Ovidius University of Constanta


206
4.3. Blended Lesson: Activities, Procedures, Timing

Note: the parts that may be transposed into blended learning are written in italics.

Table 1
Activities, Procedures, Timing

Activity Aids Interaction Procedure Time
Warm up pictures,
textbooks

PowerPoint
presentation,
web modules
PW(pair work)


e-mentoring, e-
learning
classroom, forum
T (teacher) tells students to look at
4 pictures presenting postures and
match them to the description
given.

2’


4’
Language
practice
textbooks,
body
language

flash
animation,
web modules
IW (individual
work)
CL (class)

e-learning
classroom
T tells students to use several
adjectives and the ones they already
know to speculate about sb’ s
personality, including their
colleagues.
5-10’



5’-10’
Pre-reading textbooks


PowerPoint
presentation,
text (PDF
file)
PW


e-mail, forum
T tells the students to speculate on
the meaning of the expression
“silent speech” and compare his/her
opinion with his/her colleague’s
and then read the introductory text
to see who was right.
3’


6’



Reading textbooks



text (Word
file)
IW



online
communities
T tells the students to skim the
article about body language in order
to find a possible answer to its
importance in everyday life and the
way in which one can use it in
his/her advantage.
After getting the gist, the students
are asked to read the text again for
specific information, i.e. to identify
If Clauses and some vocabulary
items, they are to use in their own
written sentences.
5’-10’



15’-20’
After-
reading
textbooks




interactive
web page
CL
IW+PW



e-learning
classroom, forum
After giving the students possible
tips for solving multiple-choice
exercises, such as the elimination
process, T gives students 10
minutes to work out the exercise on
their own and then compare their
answers in pairs.
15’




30’





The 3
rd
International Conference on Virtual Learning, ICVL 2008

207
Table 1
Activities, Procedures, Timing

Listening

tape, portable
CD/ cassette
recorder
CL
IW








PW



Students are given handouts with
exercises to be solved while
listening to the tape Big Boys Don’t
Cry, about men’s and women’
different attitude towards the same
stimuli. T asks the students to read
the statements they are to identify
as true or false while listening to the
cassette.
T asks the students to check their
answers in pairs and then discuss
the woman’s view briefly.
12’










3’
Students are announced in advance to participate in a particular class for a traditional face-to-face
activity.
Introducing
structure -
Conditionals
blackboard
handouts

























intelligent
blackboard
PW






CL








IW+PW




CL
T tells the students to imagine their
own reactions to the situations
presented on the tape and discuss
them in pairs using second and third
conditionals.

With the “help” of his/her students,
T synthesizes the structures on the
board, asking for examples and
revising introductory elements,
word order and punctuation of the
conditionals, mixed conditionals.

T asks the students to solve fill-in-
the-blanks, multiple choice
exercises and compare their
solutions in pairs.

T invites three pairs of students in
front of the class and asks them to
mime in order for the others to
identify and then imagine their
reaction in imaginary situations
such as: seeing a house on fire,
having something stolen, going to
live on a desert island.
5’






10’-15’








10’



10’
Students are announced in advance to participate in a particular class for a traditional face-to-face
activity.



University of Bucharest and Ovidius University of Constanta


208
Table 2
Anticipated Problems

Anticipated problems Possible solutions
Students may not be able to use correctly If-
clauses structures.
I will correct them and permanently
draw their attention about the structures
synthesized on the board.
T sends regularly (at a time that has been previously agreed on with his/her students)
messages containing different types of exercises, examples, extracts from films, songs) etc.


5. Conclusions

Transposing a traditional activity into blended learning (especially for
humanities) has proven to be a difficult task, because of the fact that it involves
active participation of specialists from different departments and knowledge areas,
technological support (that some students may be short of) and the time necessary
to design it. Therefore, the implementation of an E-learning programme,
following the above-mentioned steps, would prove more than useful and efficient.



REFERENCES


Books

CLARK, R. C. & MAYER, R. E. (2002), E-Learning and the Science of Instruction: Proven Guidelines
for Consumers and Designers of Multimedia Learning., Wiley & Sons, Inc. Pfeiffer, San Francisco.
DICK, W., CAREY, L. (1996), The Systematic Design of Instruction, New York, Harper Collins Publishers.
KEMP, J. E., MORRISON, G. R., & ROSS, S. M. (1996), Designing Effective Instruction, 2nd
Edition, Upper Saddle River, NJ, Prentice-Hall.
NAIDU, S. (2006), E-learning. A Guidebook of Principles, Procedures and Practices, CEMCA.
ROSENBERG, M. J. (2006), Beyond E-Learning. Pfeiffer, Approaches and Technologies to Enhance
Organizational Knowledge, Learning, and Performance, Wiley & Sons Inc. Pfeiffer, San Francisco.


Internet Sources

http://www.nwlink.com/~donclark/hrd/sat3.html
http://ed.isu.edu/addie/
http://www.personal.psu.edu/users/s/j/sjm256/portfolio/kbase/IDD/images/kempmodel.jpg
ROSSETT, A., DOUGLIS, F., FRAZEE, R. V. (2003), Strategies for Building Blended Learning,
http://www.learningcircuits.org/2003/jul2003/rossett.htm

The 3
rd
International Conference on Virtual Learning, ICVL 2008

209
Q-learning Approach in the Context of
Virtual Learning Environment

Ionita Liviu
1
, Tudor Irina
1


(1) “Petroleum-Gas” University of Ploiesti,
39 Bucharest Bd., 100680, ROMANIA
E-mail: iliviu@upg-ploiesti.ro


Abstract
Reinforcement learning (RL) is learning what to do (how to map situations to actions) to
maximize a numerical reward signal. Two characteristics: trial-and-error search
and delayed reward are the two most important features of reinforcement learning.
RL is different from supervised learning, the kind of learning studied in most current
research in machine learning, statistical pattern recognition, and artificial neural
networks. A learning problem can be solved by an intelligent agent in the context of
reinforcement learning. In this case, to obtain a lot of reward, a reinforcement
learning agent must prefer actions that it has tried in the past and found to be
effective in producing reward. In our paper we present our investigation of Q-learning
(Reinforcement Learning) in the context of Virtual Learning Environment.

Keywords: Reinforcement learning, Intelligent agent, Virtual learning.


1. Introduction

With the development of Internet technologies, online distance education is becoming
an important environment for teaching, learning, and training. In present, technologies
make it possible for more and more students to benefit from an online education.
This new paradigm in education allows students to pursue a degree without
sacrificing their jobs. Despite the advances in technology, existing online course
platforms need to deliver the highest quality online education.
The current distance education systems do not give enough consideration to many
important issues, like personalization, mobility of students and instructors, and
coordination among online study groups, which seriously affect further development of
online education.
This paper discusses the potential applications of intelligent agents to solve these
problems and make online education more accountable in the open and dynamic environment.


2. Reinforcement learning elements

Reinforcement learning is a type of machine learning and allows machines and
software agents to automatically determine the ideal behavior within a specific context, in
order to maximize its performance. Reinforcement learning is the process by which an
University of Bucharest and Ovidius University of Constanta


210
agent improves its behavior in an environment via experience. Reinforcement learning
algorithm has been widely used for many applications such as robotics, multi agent
system, game, etc. (Chen and Sycara, 1998).
Beyond the agent and the environment, a reinforcement learning system can be
described by four subelements: a policy, a reward function, a value function, and,
optionally, a model of the environment.
In the standard reinforcement-learning model, an agent is connected to its
environment via perception and action (figure 1). On each step of interaction the agent
receives as input, i, some indication of the current state, s, of the environment; the agent
then chooses an action, a, to generate as output. The action changes the state of the
environment, and the value of this state transition is communicated to the agent through a
scalar reinforcement signal, r. The agent's behavior, B, should choose actions that tend to
increase the long-run sum of values of the reinforcement signal. It can learn to do this over
time by systematic trial and error, guided by a wide variety (http://reinforcementlearning.
ai-depot.com/Tutorials.html).
In general, the model consists of the following:
• a discrete set of environment states, S;
• a discrete set of agent actions, A;
• a set of scalar reinforcement signals; typically {0,1} or the real numbers.



Figure 1. Reinforcement learning model

A policy defines the learning agent's way of behaving at a given time or is a mapping
from perceived states of the environment to actions to be taken when in those states.
Other interpretation of policy term can be associated with a set of rules. If in some cases
the policy may be a simple function or lookup table, in others it may involve extensive
computation such as a search process. The policy is the core of a reinforcement learning
agent in the sense that it alone is sufficient to determine behavior (Sutton and Barto, 1998).
A reward function defines the goal in a reinforcement learning problem. In other
words, it concentrates perceived state (or state-action pair) of the environment to a single
number, a reward, indicating the intrinsic desirability of that state. A reinforcement
learning agent's objective is to maximize the total reward it receives in the long run. The
reward function defines what the good and bad events are for the agent.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

211
A value function specifies what is good in the long run. Therefore the value of a
state is the total amount of reward an agent can expect to accumulate over the future,
starting from that state. Whereas rewards determine the immediate, intrinsic desirability of
environmental states, values indicate the long-term desirability of states after taking into
account the states that are likely to follow, and the rewards available in those states
(Sutton and Barto, 1998).
The final element of some reinforcement learning systems is a model of the
environment. Given a state and action, the model might predict the resultant next state
and next reward. Models are used for planning, meaning that any way of deciding on a
course of action by considering possible future situations before they are actually
experienced. The incorporation of models and planning into reinforcement learning
systems is a new point of view for researchers.


3. Q-learning applied in VLE

A virtual learning environment (VLE) is a software system designed to support
teaching and learning in an educational setting. A VLE allows for a course designer to
present to students, through a single, consistent, and intuitive interface, all the
components required for a course of education or training.
Universities and other institutions of higher education are increasingly turning to
VLEs in order to economize on the time of teaching staff, to provide a service for
students who increasingly look to the internet as the natural medium for finding
information and resources, to ensure the quality of teaching, to facilitate the integration of
distance learning. Many studies report the virtual learning environment as more effective,
efficient and satisfying than the traditional learning situation.
Example of VLEs: Moodle (http://moodle.org/), Claroline (http://www.claroline.net/),
ATutor (http://www.atutor.ca/). All three of these Virtual Learning Environment (VLE)
systems are written with PHP and MySQL and are based on the open source license agreement.
Extending the distance-learning platforms with agent technologies became a new
challenge for researchers and opens new possibilities in the educational prospect, complementing
the required processes of training and personalization. Agents transform distance-learning systems
from communication and information media to systems with active elements that take
part in the learning-teaching process (http://net.educause.edu/ir/library/pdf/eqm0235.pdf).
Intelligent agents can observe students interacting with educational courses, detect
learning troubles or difficulties in understanding of these students, and then suggest them
some way for overcoming those troubles (Baziukaite, 2003).
In our paper we focus on reinforcement learning approach in the context of distance
learning, giving an example of Q-learning algorithm.
Q-learning (Watkins, C., 1989) is a recent form of Reinforcement learning algorithm
that does not need a model of its environment and can be used on-line.
Q-learning algorithms works by estimating the values of state-action pairs. The
value Q(s,a) is defined to be the expected discounted sum of future rewords obtained by
taking action a from state s and following an optimal policy thereafter. Ones these values
have been learned, the optimal action from any state is the one with the highest Q-value.
After being initialized to arbitrary numbers, Q-values are estimated on the basis of
experience as following steps:
University of Bucharest and Ovidius University of Constanta


212
From the current state s, an action a is selected. This will determine a receipt of an
immediate reword r, and arrival at a next state s'.
Update Q(s,a) based upon this experience : Q(s,a) = x[r + ymaxQ(s',b)-Q(s,a)],
where x is the learning rate and 0 < y < 1 is the discount factor.
Return to the step 1.
To understand how Q-learning works we give an example as follow: suppose that
we have three pairs (state, action) as (S
i
, A
i
). The matrix R will contain the reword
function values. The row of matrix Q represents current state of the agent, the column of
matrix Q pointing to the action to go to the next state.
The transition rule of this Q learning is:

[1] )] , ( [ * ) , ( ) , ( allactions state next Q Max action state R action state Q γ + =

For our example the number of states is known (3) and the R matrix has the form:

[2]


− =
0 10
10 0
0 10 0
R .

At the beginning the Q matrix has all the elements equal to zero. The above
algorithm is used by the agent to learn from experience or training. Parameter γ has
range value of 0 to 1( 1 0 < ≤ γ ) and is called learning parameter.
As initial conditions the learning parameter has the value 0.9 and the initial state is
A2. The goal for our problem is the state A3.
For the current state A1 there are two possible actions, A2 and A3 as is presented in
the matrix R.
Now we consider that we are in state A2. There is one action corresponding to this
state, A3. The loop stops because agent reaches the goal (A3).


Figure 2. Transition process

[3] 10 0 * 9 . 0 10 )] 3 , 2 ( [ * 9 . 0 ) 2 , 1 ( ) 2 , 1 ( = + = + = A A Q Max A A R A A Q

The matrix Q will have the form:

The 3
rd
International Conference on Virtual Learning, ICVL 2008

213
[4]

=
0 0 0
0 0 0
0 10 0
Q .

By random selection a possible state of the agent could be A2 and looking in the R
matrix the possible next state is A1. From state A1 the possibilities for the agent are the
same state A1, the state A2 or the state A3.



Figure 3. The possible states for agent

[5]
9 ) 0 , 10 , 0 ( * 9 . 0 0
)] 3 , 1 ( ), 2 , 1 ( ), 1 , 1 ( [ * 9 . 0 ) 1 , 2 ( ) 1 , 2 (
= + =
= + =
Max
A A Q A A Q A A Q Max A A R A A Q


The Q matrix becomes:

[6]

=
0 0 0
0 0 9
0 10 0
Q .

When the agent reaches to the state A3 the compute process stops. Once the Q matrix
reaches almost the convergence value, the agent can reach the goal in an optimum way by
tracing the sequence of states and finding action that makes maximum Q for this state.


4. Conclusions

Intelligent agent’s applications in Virtual Learning Environment show how they
could influence teaching process. The authors give an example of Q-learning algorithm
for a teacher agent. The transition process through the set of states ends in a goal state is
presented. The best course of action is to reach the goal state with the maximum return
available.

University of Bucharest and Ovidius University of Constanta


214
REFERENCES


Books

SUTTON, S. R. and BARTO, A. G. (1998), Reinforcement Learning: An Introduction, A Bradford
Book, The MIT Press, Cambridge, Massachusetts, London, England, (http://www.cs.ualberta.
ca/~sutton/book/ebook/the-book.html)


Conference Proceedings

BAZIUKAITE, D. (2003), Concept of adaptive based virtual learning environment, in D.
Rutkauskien˙e (ed.), Proceedings of the International Conference TELDA’03, Kaunas
University of Technology, Kaunas, pp. 63-66.
CHEN, L., and SYCARA, K. (1998), Webmate: a Personal Agent for Browsing and Searching, in
Proceedings of the Second International Conference on Autonomous Agents, ACM Press,
Minnepolis, USA, pp. 132-139.


Internet Sources

http://reinforcementlearning.ai-depot.com/Tutorials.html
http://net.educause.edu/ir/library/pdf/eqm0235.pdf
http://moodle.org/
http://www.claroline.net/
http://www.atutor.ca/


Thesis

WATKINS, C. (1989), Learning from Delayed Rewards, Thesis, University of Cambridge, England.



The 3
rd
International Conference on Virtual Learning, ICVL 2008

215
Analyzing Information Security Issues
Using Data Mining Techniques

Daniela Şchiopu
1
, Irina Tudor
1

(1) “Petroleum-Gas” University of Ploiesti,
39 Bucharest Bd., 100680, Romania
E-mail: daniela_schiopu@yahoo.com


Abstract
The field of information security has grown and developed significantly in recent
years. Its concerns are: the confidentiality, integrity and availability of data with
different presentation form (electronic, print, or other forms of information). Data
mining can be applied to find relevant computer security information. In this paper
we present data mining techniques used for intrusion detection. Data provided by
National Vulnerability Database site represents the basic for our application and
our goal is to demonstrate that we can identify appropriate and accurate classifiers
to detect anomalies mediated by data mining methods. Association rules, decision
trees and other classification methods represent an effective manner for our target.
The authors aim to determine the appeared issue and to solve the information
security vulnerabilities.

Keywords: information security, vulnerability, data mining, decision tree.


1. Introduction

Networks complexity development implies the risks increase regarding networks
security. New challenges generated new strategies and solutions whose application can
provide important benefits on medium and long term. The security of a computer system
is compromised when an intrusion takes place.
To keep rules protection represents today the first priority of any user of internet
services. A network is liable to different attacks and data security became a major
objective of network administrators. To protect from such attacks, it is necessary to take
steps to prevent attacks from succeeding. In recent years various attack were reported and
a new concern appear to build viable detection and prevention attacks systems.
Data mining can be applied with success to identify intrusions and to find relevant
computer security information. In our paper we discuss about data mining methods
applied in intrusions detection as follow: association rules, decision trees, etc. At present,
there is already a multitude of various VDBs (Vulnerability Databases) containing
manifold information that can serve as basis for scientific studies. Theses databases are
driven publicly or privately by various organizations. Weka software is used in our
example and a batch of data provided by NVB (National Vulnerability Database) are data
input for our application.
University of Bucharest and Ovidius University of Constanta


216
2. Attacks and vulnerabilities types

In literature were described various types of networks attacks. In most cases the
attacks are classify in three major categories: integrity, confidentiality and availability
attacks. Researchers analyzed these networks security problems and found solutions
which appropriate implemented reduced drastically the number of attacks. The developed
systems include security politics development, users’ education and security software solution.
A system proprieties such confidentiality, integrity and availability are related.
Confidentiality and availability depend on system integrity.
Integrity property refers to data and system integrity. In first case, data integrity
means quality, correctness, authenticity and accuracy of information stored in an
informatics system. On the other hand, system integrity refers to correctively and
successfully performing of informatics resources.
Confidentiality in the framework of an IT system means that information is
available only in the conditions of security politics. System availability represents the
system property to be available only the registered users’ requests. In the context of
network security this property refers in general to the system capacity to function in the
front of denial-of-services attack.
Integrity attacks include: legalization attacks, sessions theft, content-based attacks,
protocol attacks, inattentive access methods, social engineering, unexpected information
supply, privileges fraud, confidence relations exploitation and backdoors exploitation.
Confidentiality attacks consist in: careless divulgation, information interception,
Van-Eck monitoring, hidden canals, information accumulation, etc. As availability
attacks we note: jamming, unexpected information supply, etc.
An information security vulnerability is a mistake in software that can be directly
used by a hacker to gain access to a system or network. CVE considers a mistake
vulnerability if it allows an attacker to use it to violate a reasonable security policy for
that system (this excludes excluding entirely open security policies in which all users are
trusted, or where there is no consideration of risk to the system).
An information security exposure is a system configuration issue or a mistake in
software that allows access to information or capabilities that can be used by a hacker as a
stepping-stone into a system or network.
CVE considers a configuration issue or a mistake an exposure if it does not directly
allow compromise but could be an important component of a successful attack, and is a
violation of a reasonable security policy (http://cve.mitre.org/about/ terminology.html).


3. Intrusion Detection System (IDS)


The next step in computer security after firewall was to include intrusion detection
system which try to collect information about attacks in different ways.
An intrusion detection system (IDS) monitors network traffic and monitors for
suspicious activity and alerts the system or network administrator. In some cases the IDS
The 3
rd
International Conference on Virtual Learning, ICVL 2008

217
may also respond to anomalous or malicious traffic by taking action such as blocking the
user or source IP address from accessing the network. As a definition, an intrusion
detection system (IDS) inspects all inbound and outbound network activity and identifies
suspicious patterns that may indicate a network or system attack from someone
attempting to break into or compromise a system. There are several ways to categorize
IDS: misuse detection against anomaly detection, network-based against. host- based
systems, passive system against reactive system.
In misuse detection, the IDS analyzes the information it gathers and compares it to
large databases of attack signatures. In anomaly detection, the system administrator
defines the baseline, or normal, state of the network’s traffic load, breakdown, protocol,
and typical packet size. The anomaly detector monitors network segments to compare
their state to the normal baseline and look for anomalies. On the other hand, in a network-
based system (NIDS), the individual packets flowing through a network are analyzed.
The NIDS can detect malicious packets that are designed to be overlooked by a firewall’s
simplistic filtering rules. In a host-based system, the IDS examines at the activity on each
individual computer or host. In a passive system, the IDS detects a potential security
breach, logs the information and signals an alert and in a reactive system, the IDS
responds to the suspicious activity by logging off a user or by reprogramming the firewall
to block network traffic from the suspected malicious source (http://www.webopedia.com/TERM/
I/intrusion_detection_system.html).
The most popular ISD packages are (http://www.all-internet-security.com/top
_10_IDS_software.asp): PacketAlarm, Blue Lance LT Auditor, Linux – IDS, CyberTrust
2008, Advanced Intrusion Prevention Products, Demarc Security Solution, Wireless LAN
Monitoring, Distributed Denial of Service Attacks, Honeypots – Incident Resources,
Lancope – Network Intelligence.
Intrusion detection software tends to be something of a long-term investment. It
doesn’t give instant results, but after a time the results are visible. A good intrusion
detection policy is often the one that is most strongly constructed to fit the network.


4. Data Mining Techniques as a Solution for Network Vulnerability Issue

Data mining involves the use of sophisticated data analysis tools to discover
previously unknown, valid patterns and relationships in large data sets (Two Crows
Corporation, 1999).These tools can include as well as statistical models, mathematical
algorithms, and machine learning methods (algorithms that improve their performance
automatically through experience, such as neural networks or decision trees). Data mining
consists of more than collecting and managing data; it also includes analysis and prediction.
In the last years data mining has successfully applied in domains such as
information security and our research work focused on network of excellence issues and
its vulnerabilities. Intrusion detection is defined as the process of intelligently monitoring
the events occurring in a computer system or network, analyzing them for signs of
violations of the security policy. The primary aim of Intrusion Detection Systems (IDS) is
to protect the availability, confidentiality and integrity of critical networked information
University of Bucharest and Ovidius University of Constanta


218
systems (Mukkamala, R., et al., 2000). Many applications of data mining techniques
demonstrate that information security became an interesting field for researchers witch
have the preoccupation to discover new methods to identify the most frequent issues
regarding data security in networks.
In literature, techniques such decision trees, association rules, clustering and others
are well-known in various domains and their uses for vulnerabilities classify and
clustering becomes a challenge.
On the other hand, case based reasoning approach (Lee, W. and Stolfo S., 1999)
balances with data mining techniques. Existed IDS systems are mostly static and tracks
known attacks signatures. As a result, any recognized attack is blocked from entering the
protected system and other traffic (friendly and unknown) are permitted to access the
system. Therefore malicious traffic is mostly of unknown signature type, so it will not
trigger IDS. Motivation for dynamic approach appears.
Current intrusion detection studies focuses on knowledge-based approaches that are
very efficient in detecting intruders of the type known previously, but ineffective against
new forms of threat and behaviour-based approaches, having the potential for guarding
against previously unknown types of threats. CBR can be considered as a mix of them
gathering the advantages of those approaches.


5. An Illustrative Example

In this paper we present an example of data mining techniques application in the
framework of network regarding vulnerabilities. The software used is Weka (Waikato
Environment for Knowledge Analysis) (http://www.cs.waikato.ac.nz/ml/weka/), a
collection of machine learning algorithms for data mining tasks implemented in Java
language. Weka contains tools for classification, regression, clustering, association rules,
data visualization. WEKA works with ARFF files (Attribute Relation File Format) and
also with files in .csv format (Comma Separated Values). An ARFF file is a text file
which describes a list of instances which share a lot of attributes.
The input data are taken from the NVD – National Vulnerability Database
(http://nvd.nist.gov/).
The Common Vulnerability Scoring System (CVSS) provides an open framework
for communicating the characteristics and impacts of IT vulnerabilities
(http://nvd.nist.gov/cvss.cfm). Its quantitative model ensures repeatable accurate
measurement while enabling users to see the underlying vulnerability characteristics that
were used to generate the scores. Thus, CVSS is well suited as a standard measurement
system for industries, organizations, and governments that need accurate and consistent
vulnerability impact scores. Two common uses of CVSS are prioritization of
vulnerability remediation activities and in calculating the severity of vulnerabilities
discovered on one's systems. The National Vulnerability Database (NVD) provides CVSS
scores for almost all known vulnerabilities.
The variables used in the statistics are the base score, the exploit subscore, the
impact subscore, the product name (operating system and other products) and severity
The 3
rd
International Conference on Virtual Learning, ICVL 2008

219
(gravity of the attack). The authors intend to determine how the gravity of the attacks depends
on the others four variables (base score, exploit score, impact subscore and product name).
The database used in this application is in .csv format and contains five attributes in different
data format (nominal, numeric) and 11000 records. The structure of database is presented in Figure 1.



Figure 1. The database structure (.csv format)

The database in the Viewer window in Weka is presented in Figure 2.



Figure 2. Viewer window of Weka with input data set
University of Bucharest and Ovidius University of Constanta


220
The input data are taken from NVD, from years 2003 and 2006. Weka gives us, in
preprocess step, informations about attributes, instances, several statistics (see Figure 3).



Figure 3. Preprocessing data: relation, number of instances, statistics for each attribute

In the current example, we used the Weka classifier algorithms – RandomTree and
J48 (similar with ID3 algorithm, but in this case we can’t use ID3 because we have
numeric attributtes) – to generate decision trees.
After RandomTree execution with Weka, we obtained series of results presented in
figure below:



Figure 4. RandomTree algorithm results
The 3
rd
International Conference on Virtual Learning, ICVL 2008

221
The decision tree obtained provides decision rules, some of them being presented below:
– IF Indice_de_exploatare < 9.3 AND Indice_de_baza < 3.8 THEN Gravitate = Low;
– IF Indice_de_exploatare < 9.3 AND Indice_de_baza > = 3.8 AND
Nume_produs = Windows_2000 AND Indice_de_impact < 8.45 THEN Gravitate = Medium;
– IF Indice_de_exploatare < 9.3 AND Indice_de_baza > = 3.8 AND
Nume_produs = Windows_XP AND Indice_de_impact < 8.45 THEN Gravitate = Medium;
– IF Indice_de_exploatare < 9.3 AND Indice_de_baza > = 3.8 AND
Nume_produs = Windows_Server_2003 AND Indice_de_exploatare < 4.4 AND
Indice_de_impact < 8.45 THEN Gravitate = Medium;
– IF Indice_de_exploatare < 9.3 AND Indice_de_baza > = 3.8 AND
Nume_produs = Windows_Server_2000 AND Indice_de_exploatare > = 6.25 THEN
Gravitate = High;
– IF Indice_de_exploatare < 9.3 AND Indice_de_baza > = 3.8 AND
Nume_produs = SuSE_Linux AND Indice_de_impact > = 8.45 THEN Gravitate = High;
In this case, size of tree is 1231, root mean square error is 0.0014, relative absolute
error is 0.0126% and root relative squared error is 0.3205%.
After J48 execution with Weka, we obtain the following tree with 5 nodes:



Figure 5. J48 tree

The decision tree obtained provides three decision rules, as follow:
– IF Indice_de_baza >6.8 THEN Gravitate = High;
– IF Indice_de_baza <= 3.6 THEN Gravitate = Low;
– IF Indice_de_baza > 3.6 AND Indice_de_baza <= 6.8 THEN Gravitate = Medium.
As we notice, in the second case, the pruned tree is reduced very much, so the
severity of the attacks not depend on product name, or on exploit subscore, or on impact
subscore.
Consequently, RandomTree algorithm offers a better classification for the variable
severity.


University of Bucharest and Ovidius University of Constanta


222
6. Conclusions

Intrusion detection is an important task for information infrastructure security. One
major challenge in intrusion detection is that we have to identify and classify the hidden
intrusions from a huge amount of normal communication activities.
The data mining techniques are viable solutions for determine the severity of the attacks
and they can be included in IDS, generating the development of IDS based on Data Mining.



REFERENCES


Books

Two Crows Corporation (1999), Introduction to Data Mining and Knowledge Discovery, Third
Edition (http://www.twocrows.com/intro-dm.pdf)


Book Chapters

JENKINS, D. (1983), Quality of Working Life: Trends and Directions, in H. Kolosny and H. van Beinum (eds.),
The Quality of Working Life and the 1980s, Praegner, New York.


Conference Proceedings

MUKKAMALA, R., GAGNON, J. and JAIODIA, S. (2000), Integrating Data Mining Techniques
with Intrusion Detection Methods, in Proceedings of the IFIP WG 11.3 Thirteenth
International Conference on Database Security: Research Advances in Database and
Information systems security, 33-46.
LEE, W. and STOLFO, S., “A Framework for Constructing Features and Models for Intrusion
Detection Systems”, in Proceedings of the 5th ACM SIGKDD International Conference on
Knowledge Discovery and Data Mining, San Diego, Calif, USA, August 1999.


Internet Sources

http://cve.mitre.org/about/ terminology.html
http://www.cs.waikato.ac.nz/ml/weka/
http://nvd.nist.gov/
http://nvd.nist.gov/cvss.cfm
http://www.webopedia.com/TERM/I/intrusion_detection_system.html
http://www.all-internet-security.com/top_10_IDS_software.asp


Computer Programs

The University of Waikato, New Zeeland, Weka (Waikato Environment for Knowledge Analysis) version 3.5.6.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

223
Involving Learner’s Emotional Behaviors
in Learning Process As a Temporary Learner Model

Ahmad Kardan
1
, Younes Einavypour
1


(1) Advanced e-Learning Technologies Group,
Department of Computer Engineering and Information Technology,
Amirkabir University of Technology (Tehran Polytechnic),
424, Hafez St., Tehran, 15875-4413, Iran
E-mail: {aakardan, younos}@aut.ac.ir


Abstract
It is essential for adaptive learning systems to have information about the learner.
More information about the learner provides more precise deduction about
learner’s knowledge, characteristics and preferences. This information is stored in
learner model. Furthermore emotions play an important role in cognitive processes
and therefore it must be regarded in online learning. Besides learner emotional
behaviors should be considered different from other learner's characteristics like
knowledge and preferences. They also may be different in different sessions. Regular
learner modeling focuses on prerequisite relationships for updating the learner
model. But in the example based learning, learner ability for doing exercise must be
regarded as well. In this case, learner’s emotional behaviors may lead to wrong updating
of the learner model. On the other hand, learner inability in doing exercise may be
related to his/her emotional state. Also, It may be temporary and only for this learning
session. In such cases, learner model should not be updated by temporary results until
the system ensures that they are right. In this work a new model for learner modeling will
be represented that divides learner model into two parts: 1 – permanent learner model and
2 – temporary learner model. Permanent learner model stores information about learner
knowledge and preferences, and it is utilized for other next sessions. Temporary learner
model consists of some sort of information which is only useful in the current
session. Information regarding learner emotional behaviors should be placed in the
temporary learner model of our proposed model. When the learning system ensures
that such information is not gathered from temporary emotional state of the learner,
it could be placed them in the permanent learner model. This approach leads to
more precise learner modeling for decision making.

Keywords: Learner Modeling, Temporary Learner Model, Permanent Learner Model,
Emotions Recognition, Emotional Behaviors.


1. Introduction

E-Learning systems have been progressed and provided special features which
make them more suitable for group learning. But e-learning systems are still behind face
to face tutoring by a teacher to one student (Sarrafzadeh et al., 2003). One reason for this
weakness is that they are not informed enough about learner's characteristics. In the
standard e-learning systems parameters being tracked and logged from the learner
University of Bucharest and Ovidius University of Constanta


224
behavior are few (Brusilovsky and Millán, 2007). Therefore, to enhance e-learning
systems for personalized learning, it is necessary to increase logged parameters about
learner during his/her learning process.
Another reason for this weakness is that many of the e-learning systems are not
regarding learner's emotional states (Self, 1990; De Bra and Calvi, 1998; Cheung et al.,
2003; Brusilovsky, 2003; Zhang et al., 2007). There are some previous works that have
considered to recognizing emotions of learner (Ekman, 1999; Kopecek, 2000; Sebe et al.,
2005; Osano, et al., 2006). However, they are studies focused on emotions recognition
and are not regarding enough to e-learning. For making e-learning systems more effective, we
should consider emotions of learner as well as his/her knowledge and other characteristics.
In this work we have represented a new model for learner modeling that divides
learner model into two parts: 1 – permanent learner model and 2 – temporary learner model.
Permanent learner model consists of regular learner model that stores information about
learner’s knowledge and preferences. It is usable for other sessions during the learning
process. Temporary learner model consists of some information about learner that is only
useful for this session and is not usable for other sessions. Learner emotions are placed in this part.
The organization of this paper is as follow: After introduction, in the second
section usual user modeling and emotional modeling will be described. Then, in the
section 3, our proposed model for learner modeling will be represented. This model
includes emotional model of learner in the context of temporary learner model. After that,
updating method of the permanent learner model will be explained in section 4. In this
section computational method regarding parameters like average time of solving
problems will be explained. The result of this computation is to validating this session's
gathered information. In section 5 we will discuss about evaluation of model. Finally, in
Section 6, the conclusion will be presented.


2. Usual Learner Modeling and Emotional Modeling

There are some previous works in the context of user modeling. For example,
Brosilovesky (Brosilovesky, 2003) divides adaptive hypermedia into 3 parts, which is
content model, user model and adaptation model. He is mentioned that we can store some
information about user and his/her characteristics in the user model for future usage of
them. De Bra (De Bra and Calvi, 1998) introduces user model as an overlay of content
model, which can store some parameters such as user knowledge about specific concepts
in the form of numeric values. Brosilovesky (Brosilovesky and Millán, 2007) classifies
user models of different hyper media and educational systems on the basis of used
characteristics. He mentions that Web-based adaptive educational systems (AES) mostly
rely on learner knowledge and learning goals.
There are some previous works in the context of learner emotional modeling as
well. There are 5 methods that computer could be utilized for recognition of learner's
emotional behaviors:
The 3
rd
International Conference on Virtual Learning, ICVL 2008

225
1. Questions asking,
2. Deductions making based on learner’s behaviors (Cowie et al., 2001),
3. Learner’s voice processing (Kopecek, 2000),
4. Learner’s image processing (Ekman, 1999; Pantic and Rothkrantz, 2000),
5. Learner’s behaviors monitoring using sensors (Gunes and Piccardi, 2006).
All together, since now, there hasn’t been enough consideration for learner’s
emotions in the e-learning systems.
In our proposed model, emotions are placed in the learner model as a temporary
part. Furthermore, we have recommended that newly gathered information about learner
should be placed in the temporary learner model at first. At the end of each session
system can decide which information must be moved in the permanent learner model. We
have utilized implicit parameters and asking questions for learner’s emotions recognition
in our proposed model.
Our proposed model will be discussed in the next section.


3. Our proposed model

In our proposed model, learner model is divided into two parts: Permanent Learner
Model and Temporary Learner Model (Figure 1).
Permanent learner model includes some information about learner’s previous
knowledge, his/her skills in learning, interested goals, and so on. Temporary learner
model includes logged information regarding learner's emotional attitudes during learning
process. Learner tiredness, eagerness, and so on could be tracked to describe this part of
the learner model. One of the most important issues is unexpected and short term
emotional reflections which might be affecting the temporary model. Of course, these
issues have no effect on the permanent model.
When learner starts a new session, the LMS can make use of information which is
saved in permanent learner model to make appropriate decisions for content presentation.
But all of the explanations made by the system’s reasoning should not be placed in the
permanent learner model.
Obviously, in different sessions, the environmental conditions and the learner's
emotional states may not be the same. Therefore, this difference can lead to some
mistakes in the system’s reasoning. For instance, supposing the system wants to make
decision on the basis of the number of learner’s mistakes in doing exercises. If the large
number of mistakes is took place, the system will assume that the learner himself/herself
is not prepared for starting the next chapter. But this reasoning may be caused by the
emotional conditions. Such a case is not a permanent cause and should be considered as a
temporary state. Logically, the system should not record this information as learner's
inability.
University of Bucharest and Ovidius University of Constanta


226



Figure1. Our Proposed Model

In such situations, system should compare the acquired information with the
previous ones to verify if there is any record of the same experience. If the answer is
positive, then the permanent learner model will be updated. Otherwise, when there is a
considerable difference between the acquired information and the related ones previously
saved in the permanent model, system could assume that it is necessary to interrupt
learner, and asking him/her explicitly about his/her emotional conditions.
In our proposed model, when the acquired characteristics have a large difference
with previous recorded ones (for a particular learner), they are placed in the temporary
learner model. Therefore, at the end of any session, system can decide if such information
should be used for updating the permanent learner model, or it must be ignored. In fact,
temporary learner model is like a filter (figure 2). Gathered information from this session
must be passed from this filter. Validated information will be moved to permanent learner
model for future usage. It will cause to save more precise information about learner in
permanent part of the learner model.
In the other words all system's deductions about learner are placed in the
temporary learner model at first. Then the system investigates that witch gathered
information is reasonable. Then reasonable deductions are used for updating permanent
learner model. In the next section we have explained our proposed method for updating
permanent learner model.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

227



Figure 2. Updating process of the permanent learnermodel


4. Updating the Permanent Learner Model

We have used some implicit parameters for validation of information that is saved
in the temporary learner model. The following sections illustrate how we can calculate
the value of related parameters and then, how we can use them, in the filtering process of
the learner model to develop the permanent learner model.


4.1. Implicit parameters

In this work, our focus is on two parameters: difference between mean time of
doing exercises and average of predicted mean time, and difference between percent of
occurred mistakes and average of predicted probability of making mistake in the solved
exercises.
To using our method, predicted mean time and predicted probability of making
mistake in each exercise must be saved in its meta-data. It should be predicted by a
professional teacher. In the next section we will represent our recommended equations for
calculating these parameters.


4.2. Calculations on implicit parameters

We propose equation [1] for calculating difference between mean time of doing
exercises and average of predicted mean time.

University of Bucharest and Ovidius University of Constanta


228
[1]
1 ( 1)
1
( ) ( )
1
n p n n pn
n
t t n t t
t
n
+ +
+
′ ′ − + −
∆ =
+


That
1 n
t
+
∆ is new difference between mean time of doing exercises by learner and
predicted mean time.
1 n
t
+
′ is mean time to doing exercises by learner for this session,
( 1) p n
t
+
′ is predicted mean time for this session,
n
t is previous mean time,
pn
t is previous
mean of predicted time, and n is the number of previous sessions.
Our recommended method for computing difference between percent of mistakes
and average of predicted probability of doing mistake in solved exercises is as follow. We
propose equation [2] to calculating this parameter:

[2]
1 ( 1)
1
(( ) ( ))
1
n p n n pn
n
f f n f f
f
n
+ +
+
′ ′ − + −
∆ =
+


That
1 n
f
+
∆ is new difference between percent of mistakes and average of predicted
probability of doing mistake in solved exercises.
1 n
f
+
′ is percent of learner mistakes for
this session,
( 1) p n
f
+
′ is average of predicted mistake probability for this session,
n
f is
percent of mistakes in n previous sessions,
pn
f is previous average of predicted mistake
probability, and n is the number of previous sessions. With the aid of these two
parameters, we can estimate learner’s agility in doing exercises.

[3]
( 1)
1
1 ( 1)
1
1
( )
p n
n
new
n p n
f
f
Agility m
t t
+
+
+ +
′ − ′ −
= − ×
′ ′


That
new
m is the number of exercises in new session. Then we can introduce
agility factor as follow:

[4]
1 1 ( 1) ( 1)
((1 ) (1 ) )
((1 ) (1 ) )
n n p n p n
new
n n pn pn
f t f t
m
AgilityFactor
f t f t m
+ + + +
′ ′ ′ ′ − − −
= ×
− − −


That m is average number of exercises in previous sessions. Negative or less than
α agility factor will lead to ignore this session’s results. It means that results of such a
session should not be saved in permanent learner model. We will obtain an appropriate
value for α in our future work. Obtaining an appropriate value for α will be discussed
in section 5.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

229
4.3. Filtering the Information

Regarded to learner’s unreasonable changes in time of doing exercises, percent of
mistakes, and agility, system can recognize an unsuitable emotional or environmental
state. When the system recognizes an anomaly, it can ask him/her some questions and
determine its reason. But anomalies should not be saved in the learner model
permanently, because it can make invalid conclusions. Thus, for obtaining more precise
learner modeling and to achieve better adaptation, system can save results of this session
in the temporary learner model at first, and then if no anomaly is detected it can move
them to the permanent learner model. We have named this process as Irregular
Conclusion Filtering.
If irregularity being detected is detected the system can use these unreasonable
results for adaptation in the current session, but they should be ignored at the end of the
session. For example if the number of mistakes in this session is considerably more than
other previous sessions, system can represent easier exercises to learner. But this is only a
temporary reaction. Therefore, at the end of the current session learning system can
ignore its results.


5. Model Evaluation

At present we work on evaluation of our proposed model. We have selected
mathematics course for our educational content. Our proposed model is under
implementation, and evaluation will be done on a group of 50 students at BSc degree
level. Exercises will be in various levels and predicted time for doing exercises will be
saved in the meta-data of each exercise. Then preciseness of our proposed model will be
compared with previous learner models.
Furthermore, by doing experiment we will investigate that how much each
parameter is effectible by emotions and environmental states. And then we will recognize
that what results must be detected as irregularities and must be ignored by obtaining an
appropriate value for α .


6. Conclusion

In this paper, we presented a new model for learner modeling designed for more
accurate adaptability in E-learning systems. We explained how this model could improve
the adaptability of educational environments. This improvement is feasible by dividing
learner model into two parts: 1- permanent learner model and 2- temporary learner model.
By saving irregularities in temporary learner model and ignoring them in other sessions,
we can achieve more precise learner modeling and therefore more accurate adaptability.
We are now in the implementation stage, and then the proposed models are under
examination in evaluation and validation stages.

University of Bucharest and Ovidius University of Constanta


230
REFERENCES


BRUSILOVSKY, P. (2003), Developing Adaptive Educational Hypermedia Systems: From
Design Models to Authoring Tools, in: Murray, T., Blessing, S., Ainsworth, S. (eds.):
Authoring Tools for Advanced Technology Learning Environments: Toward cost-effective
adaptive, interactive, and intelligent educational software, Dordrecht, Kluwer 377-409
BRUSILOVSKY, P., HENZE, N. (2007), Open corpus adaptive educational hypermedia, in:
Brusilovsky, P., Kobsa, A., Neidl, W. (eds.): The Adaptive Web: Methods and Strategies of
Web Personalization. Lecture Notes in Computer Science, Vol. 4321, Springer-Verlag,
Berlin Heidelberg New York.
BRUSILOVSKY, P. and MILLÁN, E., (2007), User Models for Adaptive Hypermedia and
Adaptive Educational Systems, The Adaptive Web, LNCS 4321, pp. 3-53.
CAPUS, L., and TOURIGNY, N., (2007), A Learner Model for Learning-by-Example Context
Context, Eighth ACIS International Conference on Software Engineering, Artificial
Intelligence, Networking, and Parallel/Distributed Computing.
CHEUNG,, L., B., ZHANG, H., J. and YIU, S., M., (2003), SmartTutor: An intelligent tutoring
system in web-based adult education, The Journal of Systems and Software, vol. 68, no. 1,
pp. 11-25.
COWIE, R., DOUGLAS-COWIE, E., TSAPATSOULIS, N., VOTSIS, G., KOLLIAS, S.,
FELLENZ, W., TAYLOR, J. G. (2001), Emotion Recognition in Human-Computer
Interaction, IEEE Signal Processing Magazine.
DE BRA, P., CALVI, L. (1998), AHA! An open Adaptive Hypermedia Architecture, The New
Review of Hypermedia and Multimedia 4, 115-139.
DE BRA, P. M. E. (1996), Teaching Hypertext and Hypermedia through the Web, Journal of
Universal Computer Science 2, 12, 797-804.
EKMAN, P. (1999), Facial expressions, in T. Dalgleish & T. Power (eds.), The Handbook of
Cognition and Emotion, Sussex, UK, Wiley, pp. 301-320.
GUNES, H, PICCARDI, M. (2006), Bi-Modal Emotion Recognition from Expressive Face and
Body Gestures, Journal of Network and Computer Applications, doi:10.1016/j.jnca.
09.007.
KOPECEK (2000), Emotions and Prosody in Dialogues: An Algebraic Approach Based on User
Modeling, in Proceedings of the ISCA Workshop on Speech and Emotions. Belfast, ISCA,
pp. 184-189.
OSANO, M., MARASINGHE, A., AND MADURAPPERUMA, A. (2006), A Computational
Model for Recognizing Emotion with Intensity for Machine Vision Applications, Member,
NonmembersIEICE TRANS. INF. & SYST., vol. E89-D, no. 7.
PANTIC, M., ROTHKRANTZ, L. J. M. (2000), Automatic Analysis of Facial Expressions: The
State of the Art, IEEE Trans. on Pattern Analysis and Machine Intelligence, 22(12): 1424-
1445.
SARRAFZADEH, A., GHOLAM, HOSSEINI, FAN, H., SCOTT, C., OVERMYER, P. (2003),
Facial Expression Analysis for Estimating Learner’s Emotional State in Intelligent Tutoring
Systems, in Proceedings of the The 3rd IEEE International Conference on Advanced
Learning Technologies (ICALT’03).
SEBE, N., COHEN, I., HUANG, T. S. (2005), Multimodal Emotion Recognition, Handbook of
Pattern Recognition and Computer Vision, World Scientific, ISBN 981-256-105-6.
SELF, J. A. (1990), Theoretical foundations for intelligent tutoring systems, Int. J. Artificial
Intelligence in Education 1(4), pp. 3-14.
ZHANG, Y. F., ATKINSON, R. K. & RENKL, A. (2007), Interactive Example-Based Learning
Environments: Using Interactive Elements to Encourage Effective Processing of Worked
Examples, Educ Psychol Rev 19, 375-386 doi 10.1007/s10648-007-9055-2.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

231
Classification Based on Learner’s Ability and Emotionality
For Selecting a Suitable Teaching Method

Ahmad Kardan
1
, Younes Einavypour
1


(1) Advanced e-Learning Technologies Group,
Department of Computer Engineering and Information Technology,
Amirkabir University of Technology (Tehran Polytechnic),
424, Hafez St., Tehran, 15875-4413, Iran
E-mail: {aakardan, younos}@aut.ac.ir


Abstract
Most important teacher’s duty is making learners interested. We know that it is more
important than content representation. But E-Learning systems are not regarding
enough to this reality. In the other words, they represent content in a same way to
different learners. Although adapting content to learners is good but it is not effective
enough. For more effectiveness, system could be able to adapt its tutoring method
with different learners. Thus, system could select proper method from existing
teaching methods. We have a restricted number of teaching methods in our system.
Therefore, each teaching method must be selected for a class of learners. So, each
learner, based on his/her characteristics should be placed in a correct class. System
should tutor to each class with a suitable teaching method. This method should be
designed by a psychologist team and a teacher whom is expert in educating that
content. System should be able to estimate learner classes by a little or no mistake.
Then, it should be able to adapt teaching method with different classes. In the other
words, system should behave with different classes of learners in different ways,
according to their common characteristics. In this paper we will propose an
approach for classification of learners. In our proposed method, classification is
down based on two metrics: Learner’s Rate of doing exercises and his/her
emotionality. Using these two metrics we will analyze how the system could classify
learners and how it could select appropriate teaching method for each class. We have
assigned a numeric value to each class. With our proposed method, system can
estimate classes of learners with a good probability

Keywords: E-Learning System, Learners Classification, Teaching Method.


1. Introduction

E-learning systems and educational hypermedia systems are generally consist of
three parts: learning content, learner model, and adaptation model (De Bra and Calvi,
1998; Brusilovsky, 2003). These systems are attempting to adapt educational content to
individual learners using information that is stored in learner model (De Bra, 1996;
Brusilovsky and Millán, 2007). But they don't consider enough to representing method of
this content. In the other words, their focus is on selecting an appropriate content for each
University of Bucharest and Ovidius University of Constanta


232
learner, but they don't consider to selecting an appropriate teaching method (Brusilovsky
et al., 1996; Cheung, 2003; Brusilovsky and Henze, 2007; Sonwalkar, 2007; Roy et al,
2008). It means that an effective e-learning system must represent the same content for
different learners differently. Although, there are some researches in the context of
learners classification (Daniłowicz and Kukla, 2003), but they are not consider to
important differences of learners such as learning ability and emotional affectability for
classifying them.
Different learners are differently affected by emotional motivators. In the other
words, their cognition ability is differently changeable by amending their emotional
states. Furthermore, learning ability in individual learners in the same emotional state is
not alike. We propose that e-learning system should select different teaching method for
different learners regarding to their characteristics. Characteristics that we are focused on
them are: Learning Rate and Emotional Affectability.
System can use a restricted number of teaching methods. Thus, it is essential for e-learning
system to classify the learners. Each class of learners should be related to one teaching
method. Classification is done based on three metrics that have been mentioned above.
The structure of this paper is as follow: after introduction in the second section
different classes of learners will be investigated. In the third section our proposed
methods for detecting classes of learners will be explained. Selecting an appropriate
method for representing content to each class will be discussed in the section four. Finally
in the section five evaluation results of our proposed method will be represented.


2. Learners Classes

We can classify learners in different classes for teaching them according to their
common characteristics. Learners are different in learning rate and emotionality. We can
utilize these differences for more effective representation of educational content. Each of
these two metrics has three levels: high, medium, and low (Table 1).

Table 1
Learners classes

Low Medium High
* * Learning Rate
* * Emotional Affectability

According to table 1 we can distinguish 9 classes of learners. We can assign a
numeric value (between 0 and 8) to each class. This class number must be saved in the
learner model.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

233
3. Learner class number computation method

System utilizes implicit parameters and asking some questions for class number
detection. We will explain our proposed method in the 2 next subsections.


3.1. Learner class computation according to learning rate

For learners classification based on learning rate we have focused on number of
mistakes and time of doing exercises. Our proposed system maintains probability of
doing mistake in each exercise in the meta-data of it. Also, predicted time of doing each
exercise should be stored in the meta-data of each exercise. This probability and time are
recommended by teacher of the course. We have utilized equation [1] for computing
learner's rate class.

[1]
ps
F
p
ps
1 - F
N - N
V = V - V =
T T
- V

That V V is difference between learner's rate and predicted rate.
p
V is the
predicted rate and V is learner's rate. N is the number of exercises in this session,
F
N is
the number of mistakes in this session, T is time of this session,
ps
F is average of
predicted mistake probability for exercises of this session, and
ps
T is mean of predicted
time to doing exercises in this session. If V > , > 0 a a V , learner's rate is high. If
V a b £ £ V , learner's rate is normal and if V b > , he/she belongs to low rate learners
class. The values of a and b are determined by teacher of course regarding to
educational content. Increasing the number of sessions will lead to more precise result in
learner's class detection.


3.2. Learner class computation according to learner emotionality

For detecting that how much learner is effectible according to his/her emotions,
system must estimate learner's current emotional state at first. For detecting emotions
some researchers have utilized facial or vocal recognition (Ekman, 1999; Pantic and
Rothkrantz, 2000; Gunes and Piccardi, 2006). Some others have utilized some special
sensors for movement recognition (Osano, et al., 2006). Also, there are some researches
about emotion recognizing by the other means. In this work, we request the learner to
determine his/her emotional states at the start of the session, 10 minutes later and 20
minutes later. We have assigned a numeric value for each emotional state. Positive
emotional state has positive value (+ 1) and negative one has negative value (– 1). After
University of Bucharest and Ovidius University of Constanta


234
each request sum of values is computed and after the session, average of this value is
computed as overall emotional state.
As it mentioned in previous section, mean time of doing exercises and number of
mistakes is computed for each session. System should find two sessions that learner's
emotional state value of them has most difference. Then system can use equation [2] for
detecting class of learner based on his/her emotional affectability.

[2]
2 2
1 1
2 2
1 1
3
1
2
3
1
1
ps
ps
ps
ps
i
i
i
i
F F
F F
T T
T T
CN
CN
EF





⋅ =


=
=

In the above equation, EF is learner's Emotional Factor.

=
3
1
1
i
i
CN is overall
emotional value for session by most overall emotional value.

=
3
1
2
i
i
CN is overall
emotional value for session by least overall emotional value.
1
T and
2
T are average times
of doing exercises by learner,
ps1
T and
ps2
T are average of predicted time,
1
F and
2
F are
average of mistakes for learner,
ps1
F and
ps2
F are average of predicted mistake probability
related to session 1 and 2.
If EF , 3 4 1 d d ³ £ £ , learner's affectability is low, if EF , 12 3 4 l d l £ < £ < ,
learner's affectability is medium, and if EF l < , he/she belongs to high effectible class.
Determining exact values for d and l depends on educational content and is done by the
teacher and a psychologist team.


4. Designing a teaching method for each class

In (Mayer and Allen, 1995) it has suggested that system induces learner emotions
to a suitable state. But for a learner by a little emotional affectability it could be useless.
In this case more regarding to emotional states of learner may be damage the learning
process. We propose that system behave by different classes of learners differently. We
have represented a method for dividing learners in 9 classes by means of two metrics:
learning rate and emotional affectability. We have focused on detecting classes of
learners. Designing of teaching methods is out of our discussion. For designing of
teaching methods according to each class we recommend that a psychologist team assist
the development team.
Using our proposed method we can estimate learner class by a high preciseness.
Behaving by learners according to their classes will lead to more satisfaction and thus it
can cause to more effective learning process.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

235
5. Evaluation of our proposed method

As it mentioned above, our aim is detecting of learner's class. We have assigned a
numeric value between 0 and 8 to each class. In fact, each class number is a topple for
example (High, Low). If we consider High as 2, Medium as 1, and Low as 0 then
previous topple will be (2, 0). We can represent it as 20 in ternary, and it is 6 in decimal.
In the other words, the classes are illustrated by topples (0, 0) , (0,1) , …, (2, 2) that are
correspondent by numbers 0, 1, …, 8. It means that if classes of learners are
approximately same in characteristics, their decimal values are near as well.
We will study a large number of learners to discover that how much their rate of
doing exercises are variable by occurring emotional changes. It can lead to more precise
values for d, l . Furthermore these values could be variable for different contents.


6. Conclusion

In this paper we have proposed that system should behave differently by different
classes of learners. We have classified the learners based on two metrics: learning rate
and emotional affectability. Because of emotions play an important role in the cognition
process, system should behave more carefully by emotional learners.
Moreover, we have proposed a method for detecting the class of each learner based
on our recommended metrics. We have proposed that teaching method should be adapted
by learners' characteristics for each class. Also, a teaching method should be utilized for
more than one class. Finally, we have explained the evaluation method.
Future works could be in these contexts: obtaining more precise values for a , b ,
d, l and designing a suitable teaching method for each class.



REFERENCES


BRUSILOVSKY, P. and MILLÁN, E. (2007), User Models for Adaptive Hypermedia and
Adaptive Educational Systems, The Adaptive Web, LNCS 4321, pp. 3-53.
BRUSILOVSKY, P., HENZE, N. (2007), “Open Corpus Adaptive Educational Hypermedia”, in
Brusilovsky, P., Kobsa, A., Neidl, W. (eds.), The Adaptive Web: Methods and Strategies of
Web Personalization. Lecture Notes in Computer Science, vol. 4321, Springer-Verlag, Berlin
Heidelberg New York.
BRUSILOVSKY, P. (2007), Adaptive Navigation Support, in Brusilovsky, P., Kobsa, A., Neidl,
W. (eds.), The Adaptive Web: Methods and Strategies of Web Personalization. Lecture Notes
in Computer Science, vol. 4321, Springer-Verlag, Berlin Heidelberg New York.
BRUSILOVSKY, P. (2003), Developing Adaptive Educational Hypermedia Systems: From
Design Models to Authoring Tools, in Murray, T., Blessing, S., Ainsworth, S. (eds.), Authoring
Tools for Advanced Technology Learning Environments: Toward Cost-Effective Adaptive,
Interactive, and Intelligent Educational Software, Dordrecht, Kluwer, 377-409.
University of Bucharest and Ovidius University of Constanta


236
BRUSILOVSKY, P., SCHWARZ, E., WEBER, G. (1996), ELM-ART: An Intelligent
Tutoring System on World Wide Web, in Frasson, C., Gauthier, G., Lesgold, A. (eds.),
Proc. of Third International Conference on Intelligent Tutoring Systems, ITS-96.
Lecture Notes in Computer Science, vol. 1086, Springer Verlag, 261-269.
CHEUNG, B., HUI, L., ZHANG, J. and YIU, S. M. (2003), “SmartTutor: An intelligent Tutoring System
in Web-Based Adult Education”, The Journal of Systems and Software, vol. 68, no. 1, pp. 11-25.
DE BRA, P., CALVI, L. (1998), AHA! An open Adaptive Hypermedia Architecture, The Review
of Hypermedia and Multimedia 4, 115-139.
DE BRA, P. M. E. (1996), Teaching Hypertext and Hypermedia through the Web, Journal of
Universal Computer Science 2, 12, 797-804.
DE BRA, P., RUITER, J.-P. (2001), AHA! Adaptive Hypermedia for All, in Fowler, W.,
Hasebrook, J. (eds.), Proc. of WebNet'2001, World Conference of the WWW and Internet.
AACE, 262-268.
DANIŁOWICZ, C., and KUKLA, E., The Application of Adaptive Students’ Classification to the
Determination of a Learning Strategy in an E-Learning Environment, World Transactions on
Engineering and Technology Education, 2003 UICEE.
EKMAN, P., Facial expressions, in T. Dalgleish & T. Power (eds.), The Handbook of cognitIon
and Emotion, Sussex, UK, Wiley, 1999, pp. 301-320.
GUNES, H., PICCARDI, M. (2006), Bi-Modal Emotion Recognition from Expressive Face and Body
Gestures, Journal of Network and Computer Applications, doi:10.1016/j.jnca.2006.09.007.
KOPECEK (2000), “Emotions and Prosody in Dialogues: An Algebraic Approach Based on User Modeling”,
in Proceedings of the ISCA Workshop on Speech and Emotions, Belfast, ISCA, pp. 184-189.
MAYER, J., ALLEN, J., BEAUREGARD, K. (1995), Mood Inductions for Four Specific Moods,
Journal of Mental imagery, vol. 19, 133-150.
OSANO, M., MARASINGHE, A., and MADURAPPEUMA, A. (2006), A Computational Model
for Recognizing Emotion with Intensity for Machine Vision Applications, Member,
NonmembersIEICE TRANS. INF. & SYST., vol. E89-D, no.7.
PANTIC, M., ROTHKRANTZ, L. J. M. (2000), Automatic Analysis of Facial Expressions: The State
of the Art, IEEE Trans. on Pattern Analysis and Machine Intelligence, 22(12), 1424-1445 (2000).
ROY, D., SARKAR, S. and GHOSE, S. (2008), Automatic Extraction of Pedagogic Metadata for Adaptive
Learning, International Journal of Artificial Intelligence in Education 18, 2, 97-118.
SONWALKAR, N., Adaptive Individualization (2007): The Next Generation of Online Education,
in C. Montgomerie & J. Seale (eds.), Proceedings of World Conference on Educational
Multimedia, Hypermedia and Telecommunications, pp. 3056-3063.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

237
Learning Object Tendency:
A New Concept for Adaptive Learning Improvement

Ahmad Kardan
1
, Samad Kardan
1


(1) Advanced e-Learning Technologies Group,
Computer Engineering and Information Technology Department,
Amirkabir University of Technology (Tehran Polytechnic),
424, Hafez St., Tehran, 15875-4413, IRAN
E-mail: {aakardan, skardan}@aut.ac.ir


Abstract
Improving the learning quality in e-learning environments has received
considerable attention from researchers. One of the methods to improve the
understanding of the students in a learning process is adapting the content to their
learning styles. The learning style models are used to classify the students in
different groups based on their appropriate style of learning. In e-learning we can
use the learning styles to categorize learning contents suitable for each group. In
this paper we present a new concept called Learning Object Tendency. Considering
this concept the learning objects are classified based on the learning styles of the
students. Therefore, by determining the tendency of a learning object, we can
present that the appropriate learning object to the learners .To determine the
tendency of a learning object we proposed a method based on the assessment of the
learner progress in a learning object. The Felder-Silverman Learning Style Model is
used to determine the learning style of the students. The pre-tests and post-tests are
taken before and after presentation of each learning object to estimate the level of
learning progress. By applying the results of the tests to a probabilistic model we
classify the learning object in a specific tendency.

Keywords: Learning Object Tendency, Felder-Silverman Learning Style Model,
Probabilistic Learner Knowledge Model, Adaptive E-Learning.


1. Introduction

In virtual learning environments, one of the main purposes is to minimize the
involvement of teachers during the learning process, while keeping the advantages of the
real classes in new systems.
Today, web-based virtual learning is not just presenting the learning content to
learners by means of the web. The new Intelligent Tutoring Systems (ITS) provide
intelligent adaptivity for their users. The aim of ITS is to improve the learning process of
the students without any human intervention.
Learning style is an important issue in the pedagogical physiology. This issue has
received attention from Learning Management System (LMS) developers (Mayo and
Mitrovic, 2001; Karagiannidis and Sampson, 2002; Dagger et al, 2002; Arroyo et al., 2004).
University of Bucharest and Ovidius University of Constanta


238
In this paper, we defined the “Learning Tendency” of learning Objects concept that
can be used to enhance the learning process of learners by presenting the most
appropriate learning object to them based on their learning style.
High quality learning materials are expensive to create. So it is very important to
ensure reuse of learning content. Reuse is made possible by annotating learning content
with metadata. Manual annotation is a time consuming and expensive process. It is also
liable to human errors (Roy, 2006). One of the important possible metadata for learning
content is the “Learning Tendency” introduced in this paper.
The rest of this paper is organized as follows. In section 2, we will present an
overview of the learning style models and their usage in virtual learning. Section 3
describes the structure used for scaffolding of the learning content and making a
hierarchical structure based on learning objectives, and composed of learning objects. In
the section 4, we describe the assessment methods, and the Bayesian network model that
we used for estimation of the learner’s knowledge. In the section 5, the main process of
determination of the Learning Object Tendency and the proposed Bayesian model used
for it, is explained. The section 6 demonstrates the related works in the literature. Finally,
in the section 7 the conclusion and future works is presented.


2. Learning Style Model

Different adaptations have been applied in different systems. One of the methods to
improve the understanding of the students in a learning process is adapting the content to
their learning styles. The learning style models are used to classify the students in
different groups based on their appropriate style of learning. The learning style of a
student determines what type of information he prefers, what channels he desire for
perceiving the new information, how a student processes new information and how does
he progress toward understanding.


2.1. Usage of Learning Styles in Virtual Learning

There are many different learning style models proposed for different usages. For a
list of more important learning style models, you can refer to (Karagiannidis and
Sampson, 2002). Since the population under study are engineering students, we used
Felder-Silverman Learning Style Model in this research. In the following section, we will
have a closer look at this model.


2.2. Felder-Silverman Learning Style Model

This model was proposed in 1988 by Felder and Silverman (Felder, Silverman, 1988)
for engineering students. Since then it has been used by researchers in the e-learning field
(e.g. Graf and Kinshuk, 2006; Liu et al., 2007; Sun et al., 2007). It has been revised on
The 3
rd
International Conference on Virtual Learning, ICVL 2008

239
2002 by Felder. We used the revised version that has four dimensions. These dimensions
are Sensory/Intuitive, Visual/Verbal, Active/Reflective and Sequential/Global. They address
the preferred knowledge perception, data input, processing and understanding for the
learner respectively (Felder, Silverman, 1988).


2.3. Normalization of the Learning Style Data

The proposed tool for determining the students learning style in this model is a
questionnaire called Index of Learning styles (ILS) (Felder, Soloman, 1996). To help the
students in answering this questionnaire we provided a translated version of the
questionnaire, and presented to them.
The ILS results are four numbers ranging from – 11 to 11 for each dimension. We
mapped these results to the range of 0-1 by the equation [1] where i is the ILS result for a
dimension and ns is the normalized style value.

[1]
22
11 +
=
i
ns


3. Learning Objectives and Learning Objects

Learning content scaffolding helps creating more adaptable and reusable learning
contents. Assigning a particular structure to a learning content needs some pedagogical
knowledge; this introduces the interdisciplinary research between information technology
experts and learning psychologists.
For knowledge assessment, researchers used different knowledge structures. The
Knowledge Space Theory was used by the Knowledge and Data Engineering Group of
Trinity College in their works (Conlan et al., 2006). Collins used Granularity Hierarchies,
which had been used with Bayesian belief network for Computer Adapted Testing
(Collins et a.l, 1996); in this Granularity Hierarchies, the concepts and skills are
aggregated to form levels of details.
In our research, we used the learning objective hierarchy. In this hierarchy, a
learning objective is assigned to each Learning Object. Each learning objective consists
of 1 to 3 skills. A set of questions is assigned to each skill. A question may require up to
three skills. This hierarchy is very similar to the Bayesian model shown in Figure 1.


4. Estimating learner’s progress

Determination of the “Learning Object Tendency” introduced in this paper, is based
on learner’s progress during the usage of a specific learning object. This progress is
evaluated by learner assessment before and after taking the content related to a specific
learning object (Pre-test and Post-test). Therefore, we need to accomplish a precise
University of Bucharest and Ovidius University of Constanta


240
knowledge assessment for each user. In this section, we will discuss the problems in this
process and the solution using Bayesian networks.


4.1. Learner’s Knowledge Estimation Problems

There are two common problems in assessments, when explicit tests are being used
to determine the knowledge of learners. These problems have been addressed before e.g.
(Conati et al., 2002; Pardos et al., 2006).


4.1.1. Credit-Blame Problem

This problem happens when a learner answers a question with more than one
objective being assigned to it. If the answer is not correct then, normally the blame goes
to all the skills needed for that question. Let us consider a question Q1 with two
objectives (OB1 and OB2). Prior to this question learner has answered three questions
related to OB1 and all of them are answered correctly; conversely she only answered one
of the four questions related to OB2 correctly. According to the conventional scoring system,
the blame will be divided equally between the OB1 and OB2 and the result will be:

3 of 3.5 = 86% OB1 and 1 of 4.5 = 22% OB2

However, considering the prior scores of the OB1 and OB2, it is more likely that
the learner lacked the skill of OB2 when answering this question, so the more blame must
be assigned to OB2.

Prior Scores:
¹
´
¦
= =
= =

)
`
¹


8 . 0 ) 125 / 100 ( ) 2 (
2 . 0 ) 125 / 25 ( ) 1 (

OB2 25%
OB1 100%
OB Blame
OB Blame


With this approach the scores will adjusted as follows:

3 of 3.2 = 94% OB1 and 1 of 4.8 = 21% OB2


4.1.2. Guess-Slip Problem

When a learner is faced to a test question, he/she may either have the required
knowledge for it or not. However, having the knowledge does not necessarily lead to
correct answer. On the other hand, if the question is multiple-choice he/she may give a
correct answer to a question by chance without having the required knowledge.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

241
4.2. Using Bayesian Networks to Deal With the Problems

In general, Bayesian network models have been used for assessment since mid 90s
(Martin and Vanlehn, 1995; Conati et all., 1996; Vanlehn and Martin, 1997; Conati et al.,
2002; Pardos et al., 2006). The Bayesian belief network was used for estimating the
student mastery in Computer Adapted Testing (CAT) as well (Collins et al., 1996;
Linacre, 2000; Desmarais and Pu, 2005).
The Bayesian networks have been used to address Credit-Blame in (Conati et al.,
2002). In (Pardos et al, 2006) the authors assigned a fixed chance for Guess and Slip in
answering the questions and used a Bayesian model to handle the Guess-Slip Problem.


4.2.1. Prior Probabilities

Different methods have been used to assign prior probability to nodes in the
Bayesian network model. For example, the average of former students is used as an
initiating value for the model of each new student in PKOS (Desmarais and Pu, 2005). In
a CAT algorithm suggested by Halkitis, an initial value for the ability estimate is
provided by an initialization mechanism. In this mechanism, each student is awarded one
success and one failure on two dummy questions (Linacre, 2000).
In this work, to calculate the prior probabilities we used a set of tests specifically
designed so that it requires just one skill to solve. These tests were provided to learners as
pre-tests. The scores of the tests for each skill was calculated and then mapped to a range
of 0-1. This value is assigned to the designated node for that skill in the model.


4.2.2. The Proposed Model and Methods

In order to estimate the learner’s knowledge of each Learning Objective, a
Bayesian network model as shown in Figure 1 is used. In this model, the leaf nodes
represent the questions designed for this Learning Object. To handle the credit-blame
problem, the following strategy was used:
• For the correct answers, the credit is dispatched between the parent nodes (skills)
relative to their current mastery probability.
• For each wrong answer, the blame is dispatched between the parent nodes in
reverse proportion of their current mastery probability (section 4.1.1).


University of Bucharest and Ovidius University of Constanta


242


Figure 1. The Proposed Learner Knowledge Model

For the guess-slip problem, instead of a fixed guess probability for the multiple-
choice questions (e.g. 0.25 for tests with 4 choices), a new approach was used. In this
approach, the prior probability of the parent node is also considered in calculating the
guess chance. Three levels for the prior probability of the parent node were defined (less
than 0.5, between 0.5 and 0.8 and above 0.8). In the first level, the knowledge of the
learner is considered very low, so the basic guess chance is doubled. In the second level,
the learner’s knowledge is considered low, and the basic guess chance is multiplied by
1.5. Finally, in the third level the basic guess chance is used. If the question node (Q) has
more than one parent (e.g. P1 and P2), then equation [2] is applied to calculate the
conditional probability of Q, and this conditional probability is used in the levelling system.

[2] 2 1 ) 2 , 1 | ( P P True P True P True Q p × = = = =

To adjust the slip chance, a similar levelling system is used, but this time the values
are less than 0.7, between 0.7 and 0.9 and above 0.9. In accordance, the slip chances of
0.05 for the first level, 0.1 for the second level and 0.2 for the third level are used. The
guess and slip chances are used in the conditional probability tables of the question nodes.
Table 1 shows the probability values of the question node “Question i”. It is a 5-choice
question shown in the model presented in Figure 1. The values are related to the
probability values of the node’s parents (S1 and S2). These values reflect the guess
chance for a “correct” answer and slip chance for a “wrong” answer to this question.

Table 1
Different Values of the Node Question i

Answer to the Question Node value
Correct
if (p(s1)*p(s2))>0.8 = 0.8 else if (p(s1)*p(s2))>0.5 = 0.7
otherwise = 0.6
Wrong
if (p(s1)*p(s2))>0.9 = 0.2 else if (p(s1)*p(s2))>0.5 = 0.1
otherwise = 0.05
Learning Objective
Skill 1 Skill 2 Skill 3
Question 1
Question 2
Question n Question i
Question j
Question k
The 3
rd
International Conference on Virtual Learning, ICVL 2008

243
5. Determining the Learning Object Tendency

In order to determine the Learning Object Tendency for a particular Learning
Object (LO), we compare learning progress of different learners in study of that LO,
considering their learning style. To do so, first we take a pre-test to evaluate the prior
knowledge of the learner for the learning objective related to this LO. Then we let the
learner to study the LO through the LMS. When the learner thinks that he/she has
understood the LO (and not before a determined time span, set in the LO itself) he/she
will proceed to the post-tests. We used the method and the model described in section 4 to
estimate the learner’s knowledge.
The next step is to calculate the progress of the learner. Here we used the difference
between the prior knowledge estimate (calculated statically from the pre-tests) and the
final value of knowledge estimate of the relevant Learning Objective, as the indicator for
learner progress. In this wok, we left out the negative progress, and we placed zero
progress instead.
To utilize the progress data, we found the minimum and maximum of the progress data,
and then we mapped this range to the range of 0-1. This progress data is used in the progress
nodes in the proposed Bayesian model. The other leaf node-type is the learner’s learning
style, normalized to the range of 0-1 (refer to section 2.3). After updating of the relevant
nodes, the LO Tendency is estimated as four probability values which are further
interpreted to form the LO Tendency. The proposed Bayesian model is shown in Figure
2. The tendency is determined based on the values according to the Table 2.



Figure 2. The Tendency Classification Model
LO
Sensory/Intuitive Sequential/Global Visual/Verbal Active/Reflective
Student Impact Student Impact Student Impact Student Impact
Student Progress
Student Seq/Glo Student Act/Ref Student Vis/Ver Student Sen/Int
University of Bucharest and Ovidius University of Constanta


244
6. Related Works

Automatic learning content labelling or classification is a new approach. A similar
approach is taken by Roy et al., (Roy, 2006; Roy et al., 2007; Roy et al., 2008). They
used natural language processing methods to annotate the learning content with some
predefined metadata. Compared to it our method is also a feature extraction from learning
content, but we use the experimental data for it.


7. Conclusion and Future Works

The tendency classification of learning objects can be used practically by labelling
the learning objects with the assigned tendency values; this can be added to the learning
content ontology or learning content models used in different Intelligent Learning
Management Systems to enhance the effectiveness of adaptation and improve the learning
process of the learners.

Table 2
The Tendency Classes

0-0.25 0.25-0.75 0.75-1
Sensory/Intuitive Sensory No preference Intuitive
Visual/Verbal Visual No preference Verbal
Active/Reflective Active No preference Reflective
Sequential/Global Sequential No preference Global

The Learning Object Tendency defined here can be thought as a metadata, which
makes the adaptation to learning style of learners achievable, and subsequently improves
the learning progress of the learners.
There is an undergoing research project done in the Advanced E-Learning
Technologies Lab in Amirkabir University which utilizes the methods presented here to
determine the LO Tendency for a set of contents. Another aim of this project is the
utilization of the Tendency in learning content adaptation and observation of its positive
effects in the learning progress of a virtual course.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

245
REFERENCES


ARROYO, I, BEAL, C., MURRAY, T., WALLES, R. and WOOLF, B. (2004), Wayang Outpost:
Intelligent Tutoring for High Stakes Achievement Tests, in Proceedings of the 7th
International Conference on Intelligent Tutoring Systems (ITS2004), Springer Berlin/
Heidelberg, 468-477.
CONATI, C. and VANLEHN, K. (1996), POLA: A Student Modeling Framework for Probabilistic
On-Line Assessment of Problem Solving Performance, in Proceedings of UM-96, 5th
International Conference on User Modeling, User Modeling, Inc., 75-82.
CONATI, C., GERTNER, A., VANLEHN, K. (2002), Using Bayesian Networks to manage
Uncertainty in Student Modelling, User Modeling and User-Adapted Interaction 12, 4, 371-417.
CONLAN, O., O'KEEFFE, I., HAMPSON, C. and HELLER, J. (2006), Using Knowledge Space
Theory to support Learner Modeling and Personalization, in T. Reeves and S. Yamashita
(eds.), Proceedings of World Conference on E-Learning in Corporate, Government,
Healthcare, and Higher Education 2006, Chesapeake, VA: AACE, 1912-1919.
DAGGER, D., WADE, V. and CONLAN, O. (2002), Towards a Standards-Based Approach to E-
Learning Personalization using Reusable Learning Objects, in G. Richards (ed.),
Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare,
and Higher Education 2002, Chesapeake, VA: AACE, 210-217.
DESMARAIS, M. C. and PU, X. (2005): A Bayesian Inference Adaptive Testing Framework and
its comparison with Item Response Theory, International Journal of Artificial Intelligence
in Education 15, 291-323.
FELDER, R. M. and SILVERMAN, L. K. (1988), Learning and Teaching Styles in Engineering
Education, Engineering Education 78, 7, 674-681, proceeded by a preface in 2002.
FELDER, R. M. and SOLOMAN, B. A. (1997), Index of Learning Styles Questionnaire,
http://www.engr.ncsu.edu/learningstyles/ ilsweb.html
GRAF, S. and KINSHUK (2006), An Approach for Detecting Learning Styles in Learning
Management Systems, in Proceedings of the International Conference on Advanced
Learning Technologies (ICALT 06), Kerkrade, Netherlands, 161-163.
KARAGIANNIDIS, C. and SAMPSON, D. (2002), Accommodating Learning Styles in
Adaptation Logics for Personalised Learning Systems, in P. Barker and S. Rebelsky (eds.),
Proceedings of World Conference on Educational Multimedia, Hypermedia and
Telecommunications 2002, Chesapeake, VA: AACE, 1715-1726.
LIU, F., KULJIS, J. and LINES, L. (2007), Breaking the Traditional E-Learning Mould: Support
for the Learning Preference Approach, in Human-Computer Interaction. HCI Applications
and Services, Springer Berlin/Heidelberg, 294-301.
LINACRE, J. M. (2000), Computer-Adaptive Testing: A Methodology Whose Time Has Come,
MESA Memorandum, no. 69, http://www.rasch.org/memo69.pdf
MARTIN, J. and VANLEHN, K. (1995), Student Assessment Using Bayesian Nets, International
Journal of Human-Computer Studies 42, 6, 575-591.
MAYO, M. and MITROVIC, A. (2001), Optimising ITS Behaviour with Bayesian Networks and
Decision Theory, International Journal of Artificial Intelligence in Education 12, 124-153.
PARDOS, Z. A., HEFFERNAN, N. T., ANDERSON, B. and HEFFERNAN, C. (2006), Using
Fine-Grained Skill Models to Fit Student Performance with Bayesian Networks, On-line
Proceedings of the Workshop on Educational Data Mining at the Eighth International
Conference on Intelligent Tutoring Systems, Taiwan, 5-12.
ROY, D., SARKAR S. and GHOSE, S. (2007), Learning Material Annotation for Flexible Tutorial
Systems, Journal of Intelligent Systems 16, 4, 293-305.
University of Bucharest and Ovidius University of Constanta


246
ROY, D., SARKAR S. and GHOSE, S. (2008), Automatic Extraction of Pedagogic Metadata for
Adaptive Learning, International Journal of Artificial Intelligence in Education 18, 2, 97-118.
ROY, D. (2006), Automatic Annotation of Learning Materials for E-learning, PhD Thesis: Department
of Computer Science and Engineering, Indian Institute of Technology, Kharagpur.
SUN, S., JOY, M. and GRIFFITHS, N. (2007), The Use of Learning Objects and Learning Styles
in a Multi-Agent Education System, Journal of Interactive Learning Research 18, 3, 381-398,
Chesapeake, VA: AACE.
VANLEHN, K. and MARTIN, J. (1997), Evaluation of an Assessment System Based on Bayesian
Student Modeling, International Journal of Artificial Intelligence and Education 8, 2, 179-221.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

247
Communication Models Used in the
Online Learning Environment

Gabriela Moise
1


(1) Petroleum-Gas University of Ploiesti,
no. 39 Blvd. Bucuresti, Ploiesti, ROMANIA
E-mail: gmoise@upg-ploiesti.ro


Abstract
The inter-human communication is a continuous process: a person receives,
decodes and interprets messages (according to his/her own semantics) and encodes
information in his/her answers. In 1954, Schramm and Osgood defined the circular
communication model, according to which feedback is the most important.
Communication is described as a continuous process consisting in messages sending
and receiving feedback. Starting from the circular communication model, it is
proposed a model of communication that can be used in the online learning
environment.

Keywords: multi-modal semantic communication, conversation theory, e-learning.


1. Introduction

New informational technologies permit the students to learn using computer-based
collaborative models. The computer-based collaborative techniques are relatively new.
The preoccupations of the researchers in the field of instructional design have focused on
the collaboration models among learners. In this paper it is emphasised the communication
model as support to both collaborative and unintentional learning.
“What children can do together today, they can do alone tomorrow.” (Vygotsky, 1962).
The computer-based collaborative learning does not replace the classroom
collaboration; much more it offers new opportunities to learn. The frontiers caused by
distance, cultural education, age differences, emotional states, and so forth are beyond.
Theories such as the communication theory and the conversation theory establish
expandability as a new feature of learning. The expandability of learning refers to the fact
that learning can appear anyplace, anytime, in different ways, even unintended.
The objective of this paper is to propose a model of communication that can be
used in computer-based collaborative learning combining the communication theory and
the conversational theory.


2. Communication and Conversation Theory

The communication theory was developed in 1940 in the same time with the
instruction theory. Shannon and Weaver realised an approach of the quantification and
measuring information developing the general model of communication system as
University of Bucharest and Ovidius University of Constanta


248
support of communication. (Shannon, 1948) Schramm adopted the linear model of
Shannon and Weaver to human communication, including a new concept of field of
experience, like parameter of message understanding. (Schramm, 1965) The inter-human
communication is a continuous process during which human beings encode and decode
the signals and information is encoded in answers.
The main components of online communication used in the learning process are the
context of communication, the background of knowledge, experiences, and a positive
attitude towards learning. The term of feedback in inter-human communication was
issued in the circular model of Schramm and Osgood, presented in the figure no. 1.
(Schramm, 1954) The participants on the communication process assume both the role of
transmitter and the role of receiver.





















Figure 1. The Circular Model of Inter-human Communication (Schramm and Osgood)

In essence, there are two types of communication: direct communication between
two ore more persons and mass communication.
Mass communication refers to the process of producing and freely delivering
messages to a large and heterogeneous audience. The online learning process implies a
mass communication with or without restrictions and a personalised communication. The
consequence of issuing the new communication channels is Schramm’s model of mass
communication. The Schramm’s model of mass communication is described by McQuail
and Sven, as presented in the figure no 2. (McQuail and Windahl, 1981)

Feedback
Encoder
Interpreter
Decoder
Decoder
Interpreter
Encoder

Message
Feedback
Transmitter Receiver

Message
The 3
rd
International Conference on Virtual Learning, ICVL 2008

249

















Figure 2. Mass Communication defined by Schramm

Summarizing, there are more perspectives of the communication theory:
• The technological perspective: the transmission mechanism is significant and the
effect of transmission is not essential – Shannon and Weaver model;
• The psychological perspective: the messages are filtered and processed – Schramm
and Osgood model;
• The social and cultural perspective: concerned with social interactions and
building collaboration groups.
Gordon Pask developed the conversation theory and that was used to develop
educational programmes. (Pask, 1975) The fundamental idea of the Pask’s theory is that
learning occurs through conversations about a topic. All these conversations have a
finality: knowledge building. The method of learning, according to the conversation
theory is “teachback”, i.e. a person teaches another what he/she has already learned. Pask
observed a conversation and defined the “skeleton of a conversation”.
The term used to express the learning as effect of conversation is conversational
learning. Conversational learning is viewed “as the experiential learning process as it
occurs in conversation” (Baker et all., 2002). A dialectical approach to the conversational
learning can be found in (Baker et all., 2002).
Face-to-face communication is a multi-modal process, if we take into consideration
the fact that one communicates using verbal, visual, kinesthetic expressions. In the
virtual spaces the communication must be multi-modal. The models of communication
used in the online-learning environments needs to incorporate multi-modality and
Connection at a group,
where messages are
reinterpretated and it will
act in consequence

Encoder
Interpreter
Decoder
Mass audience
A lot of receivers,
decodifications,
interpretetions,
owner codifications

Identical
messages
Inferential
feedback
University of Bucharest and Ovidius University of Constanta


250
conversational spaces. In the virtual spaces, the multi-modal communication
emulates face-to-face communications. In the paper entitled Multimodal approaches
(Project I Curriculum: The knowledge and Information skills needed for living in the
digital age http://promitheas.iacm.forth.gr/i-curriculum/), it is shown the necessity of the
multi-modal approaches in the education: “the skills needed in the digital age might
incorporate all modes that are now possible (combining voice, audio, text, images, …)
and that a new literacy approach is necessary to allow the students to use not only mode
by mode, but a combination of different modes, thus approaching outcomes to a more
natural and holistic interaction”.


3. Communication Models Used in the Online Learning Environment

The model developed by Laurillard entitled “Conversational Framework” facilities
learning as an iterative dialogue between teachers and students. The interactions
proposed in (Laurillard, 2002) operate at two levels. The former is the theoretical and
conceptual level and the latter is the practical level. This approach enables students to link
theory to practice and allows teacher to evaluate whether he/she has set or not the
adequate tasks for the learning outcomes.
The Laurillard’s model “captures the essence of university teaching as an iterative
dialogue between teacher and student(s), operating on two levels: (1) the discursive,
theoretical, conceptual level and (2) the active, practical, experiential level—the two
levels bridged by each participant engaged in the processes of adaptation (practice in
relation to theory) and reflection (theory in the light of practice).” (Laurillard, 2002)
The model developed by Salmon has five stages: stage one is named Access and
Motivation; stage two is named Online Socialization, stage three is the Information
Exchange stage, stage four is the Knowledge Construction and stage five is the
Development stage. (Salmon, 2000, 2002) The model proposed by Salmon is presented in
detail at http://www.atimod.com/e-tivities/5stage.shtml.
In this paper it is proposed a model considering the theories presented in the
previous section. The model has to integrate two perspectives: the technological
perspective and the pedagogical one. The technological perspective has three dimensions:
the human-machine communication, the machine-machine communication, and human-
human communication. The pedagogical perspective has to incorporate conversational
spaces (in this model, collaborative techniques are included in the conversational spaces).



The 3
rd
International Conference on Virtual Learning, ICVL 2008

251




















Figure 3. Perspectives of Communication in Online Learning Environment

Technological model assures the technological base of communication (hardware
and software). Some technologies that can be used to support communication are
(Rosenberg, 2006):
• E-mail;
• Mailing list;
• Discussion threads, chatrooms, forums;
• Web conferencing;
• Audio conferencing;
• Knowledge network building tools;
• News groups;
• Response pads;
• Whiteboard;
• Shared screen;
• Weblogs.
A detailed description of the way in which these technologies supports
collaboration is described in book Beyond e-Learning (Rosenberg, 2006).
In this paper, it is presented a communication model taking in consideration the
aspects from above.


Communication in the
online learning
environment

Technological model

Pedagogical model

University of Bucharest and Ovidius University of Constanta


252






































Figure 4. A Communication Model Proposed to Use in Online Learning Environment

The model presented in the figure no. 4 (Moise, 2008) has three components: the
teachers group’s component, the learner’s group component and the machine’s group
component. The forms of transfer and sharing information allowed in this model are: one-
to-one, broadcast, somebody-to-somebody and everybody-to-everybody.
identical
messages
individualized
messages
individualized
messages

identical messages

identical messages
feedback
decoder
translator
encoder
decoder
translator
encoder
decoder
translator
encoder
student
student
student
decoder
translator
encoder
decoder
translator
encoder
decoder
translator
encoder
teacher
teacher
teacher
decoder
translator
encoder
decoder
translator
encoder
decoder
translator
encoder
machine
machine
machine
individualized
messages
The 3
rd
International Conference on Virtual Learning, ICVL 2008

253
The communication is realized within each component and between components.
The messages can be personalized or identical. The reactions are obtained on the basis of
logical deductions that confer the inferential feature to the feedback.
The major problem in this model is the communication within machines group and
the communication with the machines group component. A solution to solve it is to use
the intelligent agent technology. In this case, the messages have to be more or less
“arbitrary”, more or less “standardized”. All messages have to be understood by the
participants in the communication process.
A semantic communication protocol has to be defined according to the instructional
goal of communication in online learning environment. Semantic communication can be
defined using three types of ontology: ontology of the delivered course, ontology of
online learning environment and an ontology of the e-course’s specified domain. The
model has to achieve semantic interoperability between machines and between machines
and human beings. The communication process can be both verbal and non-verbal. The
problem of interpretation of multi-modal messages is a difficult problem. In the figure no.
5 it is shown a multi-modal communication schema between two enitites. The entities A
and B can be both transmitter and receiver and the entity B is a machine.









Figure 5. Multi-modal Communication


4. Conclusions

The paper presents a multi-modal communication model that can be used in the
online learning environment. Work of the author is currently underway to explore the
paradigm of multi-modal semantic communication. The research directions are to define
an interpretation mechanism of multi-modal messages in an online learning environment.



REFERENCES


Books

LAURILLARD, D. (2002), Rethinking University Teaching, 2nd ed., London, Routledge Falmer.
MCQUAIL, D., WINDAHL, S. (1981), Communication Models for The Study of Mass Communications,
London, UK: Longman.
PASK, G. (1975), Conversation Cognition and Learning, Elsevier, Amsterdam.
Entity A
(Human being or
machine)
Entity B
(Machine)
Multi-modal
message
Multi-modal
message
Ontologies
Inference engine to
interpret
multi-modal
messages
University of Bucharest and Ovidius University of Constanta


254
ROSENBERG, M. J. (2006), Beyond E-Learning, Approaches and technologies to Enhance
Organizational Knowledge, Learning, and Performance, Pfeiffer, U.S.
SALMON, G. (2000), E-moderating: The Key to Teaching and Learning Online (2nd edition),
New York: RoutledgeFalmer.
SALMON, G. (2002), E-tivities. The Key to Active Online Learning, Taylor & Francis.
SCHRAMM, W. (1965), The Process and Effects of Mass Communication, 6th ed. Urbana, IL:
University of Illinois Press.
VYGOTSKY, L. (1962), Thought and language, Cambridge, MA: MIT Press.


Book Chapters

BAKER, A., JENSEN, P. J., & KOLB, D. A. (2002), Conversational Learning: An Experiential
Approach To Knowledge Creation, Westport, Connecticut: Quorum Books.


Journal Articles

LAURRILARD, D. (2002), Rethinking Teaching for the Knowledge Society, EDUCAUSE
Review, vol. 37, no. 1, pp. 16-25.
SHANNON, C. E. (1948), A Mathematical Theory of Communication, Bell System Technical
Journal, vol. 27, pp. 379-423, 623-656.


Theses

MOISE, G. (2008), Contributions to the Modelling and Controlling the Online Instructional
Process Using Artificial Intelligence Techniques, Petroleum-Gas University of Ploieşti.


Internet Sources

http://www.cultsock.ndirect.co.uk/MUHome/cshtml/index.html
http://promitheas.iacm.forth.gr/i-curriculum
http://www.atimod.com/e-tivities/5stage.shtml

The 3
rd
International Conference on Virtual Learning, ICVL 2008

255
Differential Geometry of Space Curves with Mathcad

Nicolae DăneŃ

Technical University of Civil Engineering of Bucharest
124, Lacul Tei Blvd., Bucharest, RO-020396, ROMANIA
E-mail: ndanet@cfdp.utcb.ro


Abstract
The Frenet trihedron is the most important topic in the differential geometry of
space curves. The paper presents an example of a lecture about the Frenet trihedron
developed using Mathcad.

Keywords: Differential geometry, Space curves, Frenet trihedron, Mathcad.


1. Introduction

In 2005/2006 academic year Romania adopted the new higher education structure
base on three cycles (Bachelor, Masters’ and Doctoral studies) according to the Bologna
Program. In technical universities the standard length of studies in the first cycle
(Bachelor’s degree) is four years. During these years students take mathematics courses
only in the first year (with some exceptions).
At the Technical University of Civil Engineering of Bucharest, Faculty of
Railways, Roads and Bridges, the Department of Mathematics and Computer Science
delivers courses only to the students in the first year. The course Linear Algebra,
Analytical and Differential Geometry is scheduled in the first semester together with
Mathematical Analysis (I) and Using Computers. In this context the differential geometry
of curves, a very important chapter of mathematics for the future designers of highways
and railways, must be taught in a very short time (two weeks). In my opinion, in this
situation a solution for increasing the students’ understanding of mathematical concepts is
to use computer technology not only for teaching and graphical visualizations but also for
solving problems.
The paper presents an example of lecture about the Frenet trihedron, one of the
most important topics in the study of space curves. Teaching this subject traditionally is a
difficult task because not every student can imagine this moving system of coordinates
attached to a curve in each of its points. Using Mathcad worksheet presented in Section 3
the teacher can quickly show the image of this trihedron and the student can use it at
home to study alone different space curves and to solve problems.
Among many others mathematical programs I chose to use Mathcad because this
software closely resembles with a worksheet. Having a well-designed graphical user
interface based on what-you-see-is-what-you-get feature, Mathcad is easy to learn and
easy to use. Equations in Mathcad look like the way we write them on the paper or on the
blackboard. In Mathcad it is easy to combine text regions, graphical representations and
University of Bucharest and Ovidius University of Constanta


256
math equations obtaining complex mathematical documents. The students are initiated to
use Mathcad during the Using Computers course in the first part of the first semester.
In Section 2 some basic facts about space curves are recalled. The aim of this
section is to establish the notations and to write the formulas used later in the Mathcad
implementation. Section 3 presents a Mathcad worksheet about how to plot the Frenet
trihedron. The reader is invited to observe carefully how the formulas from Section 2 are
written in the Mathcad worksheet. Since the plotting of a space curve and a surface in the
same drawing is not always an easy task in Mathcad, we present in detail the formatting
of the graph region at every step. Section 4 contains conclusions.


2. Space Curves. Some Background Notions

In this section we establish notation and terminology used throughout the paper and
recall the basic notions about space curves, especially about the Frenet trihedron.
By a space curve we shall understand the image of a vector-valued function

3
: R R r → ⊂ I
r
, k j i r
r r r
r
) ( ) ( ) ( ) ( t z t y t x t + + = ,
(1)

which is one-to-one and continuously differentiable on the interval I . (The mathematical
correct definition of a space curve is not the subject of this paper.) The relation (1) is
referred to as a parametric equation (representation) of the curve ) (t r
r
. The vector
k j i r
r r r
r
) ( ' ) ( ' ) ( ' ) ( ' t z t y t x t + + = is directed along the tangent to the curve at all regular
points, i.e., the points where 0 r
r
r
≠ ) ( ' t . A regular curve is a curve that has only regular
points. In what follows we shall consider only regular curves that have triply continuously
differentiable equations and satisfy the condition 0 r r
r
r r
≠ × ) ( ' ' ) ( ' t t for all I t ∈ .
At every point ) (t M on a curve having an arbitrary parameterization (i.e., the
parameter t is not the arc length) we can construct a moving trihedron, called the Frenet
trihedron or Frenet frame, formed by three mutually orthogonal unit vectors

) ( '
) ( '
) (
t
t
t
r
r
τ
r
r
r
= ,
) ( ' ' ) ( '
) ( ' ' ) ( '
) (
t t
t t
t
r r
r r
β
r r
r r
r
×
×
= , ) ( ) ( ) ( t t t τ β ν
r
r
r
× = .
(2)

The vectors ) ( ), ( t t ν τ
r r
and ) (t β
r
determine, respectively, the directions of the
straight lines called the tangent, the principal normal and the binormal to the curve at
the point ) (t M . The vectors defined in (2) are called the unit vectors of the Frenet
trihedron at the point ) (t M .
The 3
rd
International Conference on Virtual Learning, ICVL 2008

257
These three vectors ) ( ), ( t t ν τ
r r
and ) (t β
r
specify a coordinate system (for which
they are the base vectors) at each point ) (t M of the curve, the system varying as the
point moves along the curve. The axes of this coordinate system are:
1) the tangent (its direction is determined by the vector ) (t τ
r
).
2) the principal normal (which goes along the vector ) (t ν
r
).
3) the binormal (its direction coincides with that of the vector ) (t β
r
).
The coordinate planes of the system are:
1) the normal plane; it is the plane drawn through the point ) (t M
perpendicularly to ) (t τ
r
.
2) the rectifying plane; it is the plane passing through the point ) (t M
perpendicularly to ) (t ν
r
.
3) the osculating plane; it is the plane passing through the point ) (t M and
perpendicularly to ) (t β
r
.
To write the equations of these axes and planes we use their vector form. We
recalled that the vector equation of a straight line passing trough a point
0
M which has
the position vector
0
r
r
in the direction of a vector a
r
is

a r ρ
r r r
λ λ + =
0
) ( , R ∈ λ .

(Here k j i ρ
r r r
r
) ( ) ( ) ( ) ( λ λ λ λ z y x + + = denotes the position vector of an arbitrary
point on the straight line.) Then the equations of the axes determined by the unit vectors
of the Frenet trihedron to the curve ) (t r
r
at the point ) (t M are:

Tangent line ) ( ) ( ) ( t t τ r ρ
r r r
λ λ + = , R ∈ λ , (3)
Principal normal ) ( ) ( ) ( t t ν r ρ
r r r
λ λ + = , R ∈ λ , (4)
Binormal ) ( ) ( ) ( t t β r ρ
r
r r
λ λ + = , R ∈ λ , (5)

The vector equation of a plane passing trough the point
0
M , which has the position
vector
0
r
r
and is parallel with the directions of two non-collinear vectors a
r
and b
r
is

b a r ρ
r
r r r
v u v u + + =
0
) , ( ,
2
) , ( R ∈ v u .

(Here k j i ρ
r r r
r
) , ( ) , ( ) , ( ) , ( v u z v u y v u x v u + + = denotes the position vector of an
arbitrary point on the plane.) The equations of the planes determined by the axes of the
Frenet trihedron at the point ) (t M are:
University of Bucharest and Ovidius University of Constanta


258

Normal plane
) ( ) ( ) ( ) , ( t v t u t v u β ν r ρ
r
r r r
+ + = ,
2
) , ( R ∈ v u ,
(6)
Rectifying plane
) ( ) ( ) ( ) , ( t v t u t v u β τ r ρ
r
r r r
+ + = ,
2
) , ( R ∈ v u
(7)
Osculating plane
) ( ) ( ) ( ) , ( t v t u t v u ν τ r ρ
r r r r
+ + = ,
2
) , ( R ∈ v u ,
(8)

The shape of a space curve in the vicinity of its point ) (t M is characterized by two
real numbers: ) (t K , the curvature, and ) (t T , the torsion, defined below

3
) ( '
) ( ' ' ) ( '
) (
t
t t
t K
r
r r
r
r r
×
= ,
2
) ( ' ' ) ( '
) ( ' ' ' )) ( ' ' ) ( ' (
) (
t t
t t t
t T
r r
r r r
r r
r r r
×
⋅ ×
= .
(9)

For other unexplained terminology about space curves see, e.g., (Lipschutz, 1969;
Budak and Fomin, 1973; Rovenski, 2000). The readers interested in developing their
Mathcad skills for drawing curves and surfaces can consult the e-books (Lorczack, 2001;
Birkeland).


3. Frenet Trihedron. A Mathcad Implementation

In this section we provide a full Mathcad worksheet for plotting the Frenet
trihedron to a space curve in a fixed point. The space curve is defined by its vector
equation since this form permits to express the formulas, especially for curvature and
torsion (9), in a simple manner using vector calculus.
First we define the vector equation of the curve and its first tree derivatives which
are necessary for all the calculations in the entire worksheet.



To compute the derivatives we use symbolic calculation.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

259



The formulas for curvature and torsion at each point of the space curve are given below.



In order to obtain a fixed point M0 on the curve we give a value to the parameter t,
for example

t0 1 :=


Then we define the point M0 and compute symbolically its Cartesian coordinates



For graphical representation of the point M0 it is necessary to define each Cartesian
coordinate of the point thus



Then open the 3D plot operator following the path: Insert, Graph, 3-D Scatter Plot,
and complete the placeholder with (x0,y0,z0). To obtain a visible point, double-click the
graph region to open the multi-tabbed dialog box 3-D Plot Format, click on the
Appearance tab and at Point Options tab define the dots’ size of the point 4 and choose
the color red in the box to the right of Solid Color (Figure 1). The result is presented
in Figure 2.

University of Bucharest and Ovidius University of Constanta


260



Figure 1. Figure 2.

To plot the space curve type a coma after (x0,y0,z0) in the graph region and
complete the new placeholder with the name of the vector-valued function, r. Surprising,
we will obtain a surface and not a curve (Figure 3).
To obtain the plot of a curve and not of a surface, double-click on the graph to
appear again the 3-D Plot Format multi-tabbed dialog box, press the General tab for
Plot 2 and select in the tab named Display As the option Scatter Plot. Thus we obtain a
curve, but this is plotted with few points (Figure 4). To change the number of points used
for plotting the curve, double-click again the graph, press the QuickPlot Data tab for
Plot 2, and change the # of Grids from the default value 20 to a bigger one. Note that this
value must be an integer between 1 and 200. Set this value 200. The plot obtained is
showed in Figure 5.


Figure 3. Figure 4. Figure 5.

Now we define the unit vectors of the Frenet trihedron for tangent, principal
normal and binormal (see the formulas (2) in Section 2).



The 3
rd
International Conference on Virtual Learning, ICVL 2008

261
Using these vectors we define the vector equations of the tangent, principal
normal and binormal denoted by ρt, ρn and ρb in order to distinguish between them. In
Mathcad we can not have the same names for different objects. (Compare these equations
with the formulas (3), (4) and (5) from Section 2.)



To plot the tangent line we must add the name of the vector equation of the tangent
ρt in the graph region shown in Figure 5. In the first step we obtain a surface like in
Figure 6. To obtain a line double-click the graph and make the following changes in the
3-D Plot Format multi-tabbed dialog box:
(i) In General tab, Plot 3, mark Scatter Plot.
(ii) In QuickPlot Data tab, Plot 3, at # of Grids increase the number from 20 to 200.
(iii) In Appearance tab, Plot 3, at Point Options, choose Solid Color Blue.
See the result in Figure 7.
Proceed analogously for principal normal and binormal. Choose different colors for
different lines. For example, plot the principal normal with magenta and the binormal
with green. Figure 8 shows all the axes of the Frenet trihedron at the point M0.



Figure 6. Figure 7. Figure 8.

Now we define the vector equations of the planes determined by the unit vectors
ν τ
r r
, and β
r
of the Frenet trihedron at the point M0. We use the notation PN for normal
plane, PR for rectifying plane and PO for osculating plane. In Mathcad it is mandatory
University of Bucharest and Ovidius University of Constanta


262
to have different names for different planes. Compare the formulas (6), (7) and (8) of
Section 2 with the formulas below. Evaluating symbolically every vector equation we
show which are the Cartesian components of every plane.




To draw the normal plane double click the graph and add the name of the vector
equation of the normal plane, PN. Then double-click again the graph and make the
following changes in the 3-D Plot Format multi-tabbed dialog box:
(i) In General tab, Plot 6, mark Surface Plot.
(ii) In Appearance tab, Plot 6, at Fill Options, mark No Fill, and
(iii) At Line Options choose Wireframe and at Solid Color choose Light Blue.
The result is showed in Figure 9. Proceed in the same way for the rectifying plane
(Figure 10) and for the osculating plane (Figure 11) choosing for each of them a light
color in concordance with the color of the unit vector perpendicular on the corresponding
plane (magenta, respectively green).


Figure 9. Figure 10. Figure 11.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

263
Finally, we put together all the axes and the planes in the same drawing. The Frenet
trihedron to the given curve at the point M0 is showed in Figure 12.


Figure 12.

We end the worksheet by computing the curvature and the torsion of the curve at
the point M0.




4. Conclusions

The differential geometry of space curves is an important topic for civil engineering
students, especially if they will be the future designers of highways and railways. Unfortunately,
there is insufficient time to teach this chapter together with linear algebra and analytical
geometry in the first semester. A solution can be the use of computer technology, and this
paper shows how to use Mathcad to visualize the Frenet trihedron to a space curve.
Using the Mathcad worksheet presented in Section 3 the students learn to plot
space curves and points on these, to relate graphical objects to their analytic definitions,
and to see the graphical effects of varying parameters. The use of Mathcad can help
students to develop their mathematical skills and to deeper understand the theoretical
concepts. The Mathcad capability to do symbolic calculations allows students to focus
their attention on understanding mathematics concepts and not on hard computations.



REFERENCES


BIRKELAND, B., Creating Amazing Images with Mathcad, Mathcad E-Book; http://www.ptc.com/
appserver/mkt/products/resource/mathcad/
University of Bucharest and Ovidius University of Constanta


264
BUDAK, B. M. and FOMIN, S. V. (1973), Multiple Integrals, Field Theory and Series. An
Advanced Course in Higher Mathematics, Mir Publishers, Moscow.
LIPSCHUTZ, M. M. (1969), Differential Geometry, Schaum’s Outline Series, McGraw-Hill, New York,
San Francisco.
LORCZAK, P. R. (2001), 3D Plotting from the Mathcad Treasury. Updated to Mathcad 2001, MathSoft
Engineering and Education, Inc.
ROVENSKI, V. (2000), Geometry of Curves and Surfaces with MAPLE, Birlhäuser, Boston,
Basel, Berlin.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

265
Safety in Web 2.0

Dumitrescu Marina
1
, Dumitrescu Bogdan
2
, Daniel Raduta
3


(1) Technical College of Telecommunication „Nicolae Vasilescu-Karpen”,
Bacau, mimozad2005@yahoo.com
(2) Dumitrescu Bogdan – University of Bucharest,
bogdumdum@yahoo.com
(3) Daniel Raduta – Security Department of Bitdefender AntiSpam,
Softwin – Bucharest, draduta@yahoo.com


Abstract
The Internet is no longer a tool for information as it was in the past. Now it has it’s own
personality, complexity and raises serious problems to those who under estimate it.

Keywords: BlackHat, Manipulation, Social Networks.


1. Introduction

In a study conducted by the Department of Anthropology, University of
California, Los Angeles, social sciences researcher Peter J. Richerson tried to bring on
the critical list, the man and his interraction with the multi-media phenomena. He started
from the assumption that "unlike other organisms, the man receives information in a
complex format, from others through social learning techniques such as imitation. This
information is captured by human in a consciously maneer or subconsciously level, a
level of the human behavior."

“A behavior is always to be taken transactionally: ie., never as of the organism
alone, any more than of the environment alone, but always as of the organic-environmental
situation, with organisms and environmental objects taken as equally its aspect. “ (Dewey
si Bentley, 1949)


1.1. Web 2.0

The concept of Web 2.0 was born along with the beginning of new era of the
Internet market when intelligent technologies concerning advanced online programming
were developed, technologies like Ajax and Php5, technologies that together with the
advantages offered by the Java platform revolutioned this field of information. The pages
are not only sources of information in an environment free of connections just like in the
beginning, now they have a high degree of interactivity thanks to the Ajax system
because now they can identify and run separate and personalized processes.
University of Bucharest and Ovidius University of Constanta


266
Gaining this degree of freedom, Web 2.0 offers a number of advantages and
disadvantages, allowing on one hand, interaction with the human user, but also leaving
access to certain sensitive information placed on the machine had in posession.


1.2. Psychologic complexity

Since the XXI generation, multimedia resources available for masses, determined
an accelerated evolution of the knowledge and now, humans can capture information
starting from an early increasingly age. The society, as it is today and especially in the
industrialised countries, has developed a wide range of multi-media industry to conquer
all areas of the market and to "catch" all categories of age in this game and trend of the so
called „new generation”.
0
20
40
60
80
100
2002 2003 2004 2005 2006 2007
General Traffic
Traffic on social
networks
Traffic on grey
locations

Chart 1. Traffic from 2002 to 2007

A series of statistics provided by Google Analytics conducted on the large engines
of social networks reveal a colosal traffic from the highly developed countries on a
comprehensive range of services, which covers 90% of the current topics. Statistics made
in accordance with ISP’s and a series of market analysis, reveals that the social networks
are filled preponderently with children and young people in search of new.
The graph shows how, since 2002 when the phenomena was registered for the first
time, and the evolution of this, the traffic on social networks and locations grey had been
alarmingly increasing from year to year. This represents an average graphic report of all
countries surveyed, but only those under or in developing and developed ones.


1.3. Tipization

As time flows, conception and human mentality had gradually diversified. Current
technology allows "tasting" all the sensations offered by life and the freedom to choose
his own lifestyle. Unlike the media channels available for the population 20 years ago,
The 3
rd
International Conference on Virtual Learning, ICVL 2008

267
when all the content and the ideas were censored, the Internet was used in small circles
and in certain limits, now this has been exceeded, the Internet is a fully free medium and
without limited data areas.
The only limits are those that the user defines, but they can be easily avoided, by
techniques of social engineering. If 20 years ago the society offered individual's a
lifestyle, a mentality and unity, now inflow data allows them to leave this area and choose
a unique path in life. We can no longer talk in the 21st century about a mass of people
who can be included in a pattern.


1.4. Patterns

Although the paths chosen by individuals, allow them to be different in all aspects
of life, analyses, however, from the dedicated deparment in San Francisco University
which is studying human behavior (focused especially on Informatics), states that they
still can be classified. Regardless of religion, social status, even listened music and
conception of life, teenagers have certain features that identify them in any situation,
features which handled in a proper manner can be used for handling masses.


2. Psichological reports

“Unlike other organisms, humans acquire a rich body of information from others by
teaching, imitation, and other forms of social learning, and this culturally transmitted
information strongly influences human behavior. Culture is an essential part of the human
adaptation, and as much a part of human biology as bipedal locomotion or thick enamel on
our molars. My research is focused on the evolutionary psychology of the mechanisms that
give rise to and shape human culture, and how these mechanisms interact with population
dynamic processes to shape human cultural variation. I have done much of this work in
collaboration with Peter J. Richerson.” (Dept. of Anthropology University of California,
Los Angeles, CA 90095)

Psychological reports carried out on students in last years reveal that they are
exposed to the greatest dangers they society may "provide" since their very early age.
This is materialized, in various "attacks" on the part of phisics, when a person tries to
have a direct contact with them, either through derived methods using the computers.
Internet through the Web 2.0 concept that tries to promote it now, throught its
social structures in full development provides an environment for both the victim as well
as for the hunter, to get in touch very easily. The solutions available to the public by the
promoters of social media do not offer any protection and easely makes victims.
In a study conducted by the University of Los Angeles on peoples, peoples
classified by age, region of the world and levels of education reveals the deficiencies of
each category.

University of Bucharest and Ovidius University of Constanta


268
2.1. Classified by ages

Study shows that predisposal to danger is primarily influenced by the age of the
victim. Age classes for study were grouped into 3 categories: children between 10-14 years,
young people between 14 to 18 years and adults from 18 -21 years. There were no reports
and statistics made on subjects which ages cross this limits because they no longer
represent a major danger for the Web 2.0.
On the categories studied we observe that the first class, between 10 and 14 years,
has a primitive behavior, a point of view and understanding of environmental information
very very low. Therefore, at this age the child does not understand the exact nature of the
attacks and dangers and he asimilates them as they are, direct as well as the discrete,
attacks provided by the informatics network.
He is vulnerable from the simplest attacks called SPIM (SPAM on protocols as IM
[see ymsgr and gmail]) which attract the child through messages "tempting" to access or
to disclose certain information contained in a confidential area, or a more complex attack
which involves creating of an environment and with advanced solutions to manipulate the
child during longer periods.
Problems of this kind are likely in the mind of a child to change optics about life
and to initiate a new lifestyle, or waiver certain "procedures" needed only on the basis of
mentalities induced.
In terms of mental development of children, this is the period when the trends
dictate his life and any attempt to over-write them has a very high succes rate.
His neurological center acts, at this age, as a sponge and if it is exposed at regular
time to harmful information, it may have iremediabile changes in behavior, not only
changes of the perspective, but conception, a change is aspirations and mentality.
Between the ages of 14 and 18 years we meet other specific problems. The
specialized education appears, conducted by schools and colleges where teachers usually
raise alarm signals to the potential dangers to students, about the iminent dangers. The
signs are more aware of the cause, if the teacher has children at home.
During hours, an eye endorsed, may decelate, among a group of students, any type
of behaviour 'addiction' which then, together with family, class master, teachers or
psihopedagogs should try to correct them. The level of training and interaction with the
personal development of student and the institution is in a inversely proportional report to
the development of culture.
In countries in development the system tries to maintain the "subject" between
certain well-defined limits, but in industrialized and developed countries, they avoid this
behavior to "allow freedom of expression".
Researchers from San Francisco confirm the reverse proportionality, but they put
the blame on the conceptions and mentalities unanimously adopted by them.
Social networks generally refers to this category of people because they are quite
mature in the eyes of society to have certain rights to confidential data of the family, but
still in the phase in which the "boom" generated by a hoax to be perceived and treated
with a seriously enough trust to achieve the final goals.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

269

0
10
20
30
40
50
60
70
80
90
Under developed In development Developed
Under surveilence
Deviated cases

Chart 2. Analyzed cases

Once major (between 18 and 21 years) and a well-defined personality they begin to
have dreams and aspirations. It no longer represents a potential victim of social networks
because they experience life and social environment they have educated and prevented
from these dangers. Now it is attacked in other means well defined and discrete that are
interested in various organization, organizations as Blackhat and are motivated by various
advantages of a material to start their activity in this area. They use a series of social
engineering methods and hoax sites to initiate and feed the mentality of the young a a
new concept then, by means of social engineering offered the Web 2.0 is exploited. The
miraje created by these BlackHat companies has a large and deep impact among young
people because they are the result of lengthy analysis of traffic and therefore were
perfectly moulded on the psychological profiles targeted.
0
10
20
30
40
50
2005 2006 2007 2008
Hoax
Social
Engineering
BlackHat

Chart 3. Manipulation ratio

University of Bucharest and Ovidius University of Constanta


270
2.2. On regions

In the developing countries which posses new technology, but still do not have a
maturity of these services and have a specific culture to see that young children are
tempted by various tricks offered in advantageous conditions. After some polls, it was
found that the information called "tentations" spread across all channels of information
used. Due to the advance that it promotes the Web 2.0 concept this information and
tentations "can be easily extracted from specialized services to make available systems
like" trends "and as Google trends.
In under-developed countries, the tentation level is not at an alarming level. These
countries do not represent a "good" in the eyes of evil-intentioned people and therefore
did not promote specific content, and when I say here are the specific reference to a
language and a culture that are not interesting on the Internet, because the percentage of
young owners of a connection to study the conditions proved to be 8-9%. In general, and
where there is a proportion so small, they have a management-level content filters
dangers humans with a high performance automated security systems.
In developed countries, the problem is more pregnant and a more complex search
for solutions is initiated to stop the spread of the problem. Hereby young children provide
a system mature and complex that it provides all the advantages and benefits included
with the system Web 2.0. IT systems can not cover and block all the traffic and therefore
harmful actions can commit due to a lack of human operators to support the problem.
Everything is left to the machines which often have major leaks in security.
Highly evolved countries and in evolution thus companies starting to produce
products which are integrated in the standard Web 2.0 and growing as a consequence of that
fact occur overnight a number of third party products that do not benefit from an advanced
test and a security system inside and it can be used fraudulently for the purpose Blackhat.


2.3. Societies and education levels

On education levels we can distinguish:
With a high level of education, for which the computer is a common medium of
communication with friends and everyone is just like leaving the daily routine. It is
predisposed to attacks of any kind, because culture and the social status gives them the
feeling of invulnerability and do not consider any possible threat in a virtual environment.
Real life cannot provide the necessary amount of adrenaline and therefor threy are trying
to gain some by playing a game extremely violent, with an index HSRB over 10 points or
try to integrate into Blackhat networks for a new experience.
With an average level of education, the tipe that takes part in various games with
friends and finds these parties as the only "out" with friends. It is very confident in their
The 3
rd
International Conference on Virtual Learning, ICVL 2008

271
power and gives them a credibility beyond the normal limit. He lives the life between
these walls and it doesn’t have a wider vision. It is a fan of virtual social environment and
through it is manipulated. This represents an environmental incubator ideal for the most
rudimentary mass attacks.
With a low level of education, generally those with a low intellectual level and
therefore are not predisposed to attacks in the range of Web 2.0. They are marginalized
by the society and do not use social services like MySpace or FaceBook because the don’t
find identity on the internet as being important. They are users of computers mechanised
by other people from outside and defended by any problems that may arise in case of
DSET (Distributed Social Engineering Technics)


3. Fighting methods

These behavior considered derivatives under analysis and expertise on a sample of
subjects, modularized and can be programmed so that a program that integrates a network
neuronale be able to learn and to detect any problems.
The solution built on the theoretical level and raised at a global level oriented
individual tries to monitor all quantifiable factors using various algorithms. If in the case
of applications for Web 2.0 type they are supported by advanced technologies like AJAX
who can make an identification and reporting at the server side, the problem appears to
applications which are not yet integrated, or the concept does not allow them to reach
Web 2.0 techologies, therefor no sensors are included.
Here at the client side, where particularities are more problematic, trying to
implement and program a neuronal compared to monitor the subject of a long period of
time and report anomalies that appear later. Since subjects can not be predetermined, the
network must be trained before to enhance learning so that the product prove it’s
performance.
The sensors used in the client side should be integrated into an application that will
operate on a rating system and case a complex decision, which will follow the application
implementation as following:
The rate analysis of the topic for a well-defined content, which is measurable in the
virtual environment, should be analyzed in:
– In a normal time for loading a page will be marked with an X
– Abnormal conditions of time will be x, considerably lower because of the psychological
point of view has installed a "blood rush" to induce a state of aggitation. So x <<X.
Reliability of the introduction of data is another factor in the analysis, which can cause
under review, with an relative accuracy, depending on the type of depression they suffer topic:
– In this case of normality they enter data with a accuracy of X, without
generating excessive typo sites.
University of Bucharest and Ovidius University of Constanta


272
– Once installed the agitation stage, the agility of typing fails, decreased attention
and appears a number of typo’s. They will trigger an alarm signal. And here x <<X.
Tentative of disabling or avoid, in the case to take to meet the programs' parental
control "in order to hideing the osesion or problem faced.
– In case the user is during the daily routine he does not have a good reason to
stop the correct functioning of the monitoring activity being in accordance with the provisions.
– If a problem occurred, the user will try in the first phase to hide this fact and try
to solve it without raising an alarm signal to people around later instalandusing obsession.
As for the subject previously with the interruption of activities like "parental control",
the user with a medium level to advanced knowledge in science, will try to clean traces
left in the programs, files, "Temporary Files" which could indicate the locations of traffic.
Thus, this educational environment will contain the following techniques:
– Seeking class-level behavior of a unitary and pulling an alarm signal when it appears
in the behaviour of a derivative, derivative calculated by other criteria. If found a very
large deviations from the previous program begins registration of an advanced behavior in
the courts where he has made and draws a final status report to solutioning by a teacher.
A second customization of the program would be the one when it would be used in the
household. In these circumstances, the program will have included the following functions:
– Tracking computer use with or without an active program of parental outside
structures learned. This is noticeble when obsessions appear and when it is opened after
midnight, behavior which in the case of normality is not met.


4. Monitor application

The program produces a series of statistics and identify potential deviations in the
users behavior. The following table represents an analysis on a subject that during
11.03.2008 and 21.03.2008 when he had a deviated behavior.
The recorded percentage and absolute values of all methods of analysis and
monitoring program and automatically pointed out that the average value was recorded
during this period and that was the value that emerged from the pattern formed.
It can be seen that on the date 16.03.2008 the user has presented a nervous
behavior, with a considerabile number of typo’s and a high number, to precipitated by
clicks per page. Also, he presented a deviation to sites identified as having a violent
content. We recorded an increase in the use of instant mesajerie protocols, as well as
social networks. Also, using the protocol http recorded a growth, which indicates an acute
need for information, mostly violent. Although very used, the http protocol, the pages
loaded were not read in full. If he previously spent over a period of 1 to 2 minutes for
each page resulting a maximum rate of 40%, it now reaches 90%, meaning that for each
page time to read it was less than 30 seconds.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

273
Table 1
Complete analysis on a subject



They see an amplifier and traffic networks on the social and instant messaging.
Time spent on them is high traffic servers designed to mention is massive and suddenly
create a series of new relationships on IM.



Result 1. Monitor analysis

Also notice that is leaving the computer use patterns in the days of the week, both
those working as well as the weekend.
The application also provides a report in which notes the steps we face in starting
and stopping monitoring services based on user behavior. It will be submitted to a
monitoring report drawn up on the same time as the above.
University of Bucharest and Ovidius University of Constanta


274
This report analyses and determines stating with the first monitoring algorithms
used and the values they obtained during the analysis and interpretation and other acts run
through algorithms to detect in order to establish the exact behavior. It is observed that
for every program to detect it displays the normal/average as well as the value at which it
made the notification of a diversion, and is subsequently registered and action taken.



BIBLIOGRAPHY


JONATHAN BERGMAN (2006), The analysis of the modern human, Deparment of Anthrolopology,
Los Angeles.
Virus Bulletin (2008), Distributed Social Engineering Technics, Virus Bulletin.
MCGRAW HILL (2007), Web 2.0 Exposed, McGraw, iSEC.



The 3
rd
International Conference on Virtual Learning, ICVL 2008

275
Experimental Testing of Decisional

Carmen Dumitrescu
1


(1) Departament of Naval Electrotechnics, Electronics and Computers ScienceMaritime
University of Constantza,
104, Mircea Cel Bătrân Street, Constantza, Romania
E-mail: cdumit2002@yahoo.com


Abstract
The work presents the results of the scientific experimental research to design and
experiment a new procedure for computer-aided testing of maritime students
regarding their decision making capacity under crisis conditions. A software
implemented testor is described, which is meant to the experimental testing of
decisional behaviour and the experimental outcome

Keywords: computer-aided testing, decisions, crisis, merchant marine.


1. Introduction – Scope and Stages of Reasearch

Training students in mechanical engineering is performed on real ships and under
real navigation conditions. At present, testing the behaviour of maritime students under
crisis conditions is performed in labs where there is no information on actions of random
factors over their decision making. Under these circumstances some native peculiarities,
skills or talents to conceive the most proper action and behavioural tactics are not highlighted.
The aim of this scientific research is to perform an experimental study of achieving
skills and behavioural automatism under normal and crisis conditions. It is obvious that
any disturbance may influence on the training process (knowledge acquiring and
achieving skills and automatism concerning handling or running electromechanical
machinery) To be able to assess the influences and have the possibility to divide students
from the point of view of their capacity to adapt to stressful conditions, we have
organized a simulated scientific experiment, meant to test the student’s behaviour when
making decisions under crisis conditions.
According to Skinner, a stimulus is essential in the training process, and in
computer-aided testing it becomes even more important since in tutorial training the
presence and involvment of the tutor is a stimulus in itself. In examples shown by Skinner
for computer-aided training as checking procedure it is proposed to request answers built
by the student himself. Skinner has considered the action of building a correct answer a
stimulus. That it why when the student has acquired a certain amount of knowledge, and
is already studying for the following amount, it was condidered necessary that he must
have access to the correct answer for the previous amount (so as to be able to compare it
with his own). Skinner inspired himself from his experiments with pigeons. In real life
University of Bucharest and Ovidius University of Constanta


276
people train themsleves and aquire conditioned reflexes based on stimuli. This time the
stimuli are live confirmations of reaching the target within these actions. In real cases, the
possibility to achieve stimuli is at random, and when there is a possibility to achieve
them, their probability is subunit. Experimental research performed at CMU has had as
objective a procedure of testing the orientation capacity of seagoing students under such
circumstances and to what degree can their system of conditioned reflexes adapt to allow
them to take certain action and behaviour tactics so that under given crisis conditions
they may continue their existence. The experiment is under development and its first
outcome is shown in the present work.
The testor is a software meant to develop and assess the aptitudes to take decisions
under crisis conditions and under random changes of the environment, respectively. The
testor allows two consecutive stages:
• In the first stage the student is supposed to perform one of two actions A1 and A2
under different stimulus probabilities; only students who have acquired a certain assessment
quota for the first stage of binary decision test, are then promoted to the second;
• In the second stage the student directs a submersible ship for the research of
marine environment which has to avoid an obstacle; the crisis conditions in this case are
represented by the sonar indications which are jammed at random.


2. Model to Confront the Decident and the Environment

In stage I the tested student can use two identical buttons, b1 and b2, by means of
which he has to express his decision u = (b1 = 1, b2 = 0) or u = (b1 = 0, b2 = 1) pressing
one of them to determine an increase of the bargraph height S of“success” by q, when q = 1
is variabile S becomes the content of a simple counter of successful pressings:

S = S + 1, (1)

Also, the successful pressing determines a decrease by Q/T of height C of another
bargraph which signifies the cost of crisis condition:

C = C – Q/T, (2)

where T is the time interval from the last success.
The decident student has to decrease C as much as possible (by repeated binary
decisions, pressing b1 or b2 intuitively) in conditions when the effect of the decision is
affected by the hostile disturbing action of the environment. This is in fact the model of a
game between two players: the decident student and the hostile environment. The
decident transmitting + 1 on one button intends through his attempts to decrease the cost
of crisis condition fighting against the other player who is the hostile environment. The
environment opposes trying to transmit – 1 on the same button the decident is
transmitting + 1 by pressing it. The environment intends to maintain the cost constant to
The 3
rd
International Conference on Virtual Learning, ICVL 2008

277
anihilate the success in case of some of the decidents actions. Also, the decident is trying to
“avoid” the button through which the environment is transmitting at the same time s = – 1,
so that the environment may waste “ in vain” its shot by – 1. But, if the decident (by + 1)
and the environment (by – 1 ) press the same button simulataneously, the result of the
decident action is FAILURE (i.e. nil). The decident gains only if the environment
presses by – 1 a different button from the one pressed by the environment.
For the reasons afore presented the two buttons b1 and b2, available for the
decident to press, one causes at output + 1, i.e. success, with different probabilities, p1
and p2, respectively:
• Repeated pressing, N times of b1 brings about success only for a percentage of
approximately F1 of pressings,

F1 = N xP1 , (3)

where F1 is the frequency of success, and P1 is the probability to be successful when pressing
b1 (for N of hundred order the probability is approximated by relative frequency F1/N);
• Repeated pressing of N times of b2 brings about success only for a percentage of
approximately F2,

F2 = N xP2. (4)

The difference between the two players resides in the fact that the decident is able
to think and thus he may assess the characterisitcs of the tactics adopted by the
environment in the fight between the players. This way, he can adapt his own strategy so
that the percentage of success may be higher than that of failure. For example, if the
decident finds out that at a certain moment he can be successful more often by pressing,
b1 than by pressing b2, he will press only b1 from then on because it is clear that p1 > p2.


3. How the Experiment Works

The tested student does not know anything about the way the experiment is
organised and controlled by the computer. He only sees two buttons and expects the
target to appear. His mission is that when the target comes in view he should press one of
the buttons to destroy the target and make it disappear, by trying to get as many successes
of this kind. Destroying a target which comes into view is a positive stimulus in
searching for the solution which leads to a greater success. On the other way round, by
missing the target (failure l) is a negative stimulus, a sort of penalty for the bad decision,
which lead to failure. To stimulate the interest of the tested student, so as to get as good
as possible results, during the entire experiment he was informed right from the beginning
that he was competing against the computer and to defeat the computer-opponent he had
to mobilise in order to destroy as many targets as possible. This was all the tested student
had to know and this was the only information given to the tested student before the
University of Bucharest and Ovidius University of Constanta


278
beginning of the experiment. In the first stage concerning the test of decision making
capacity in crisis conditions, the experiment was performed in two variants.
Main restrictions imposed and obeyed during the organization and performing of
the experiemnt in the first stage regarding the testing of capacity to adopt decisions under
stress conditions are shown in figure 1. The fact that the tested subject does not know
details about the experiment orgnization makes him concentrate only on the direction of
reaching the only objective, i. e. getting maximum of success.
At the end of the experiment the student has to fill in a questionnaire to explain the
way he approached decision making and how he chose the tactic to counterpart the
actions of the computer opponent.



Figure 1. Organizational restrictions for the experiment
of decision making capacity of students

As we have already shown during a testing session about N = 250 total pressings are
done either on the maximum probability button or on the minimum probability button. Of
the total N button pressings only Ns < N brings about success, and the rest of Ne = N – Ns
lead to failures. In the first testing sessions the probability of success of a button was very
high (almost 1). Under these circumstances, the subject finds out the button of maximum
success after several attempts. This was considered simple in relation to the conditions
when the two probabilities had close values. The characteristic of the experiments
performed within the second stage consists in the fact that during the entire testing
session Pmax and Pmin no longer remain attached to the same buttons during the enitre
testing session but from time to time they are reversed.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

279



Figure 2. Example on the evolution of pressings during a session

In the first stage the main characteristic of the experiment resides in the fact that
the values Pmax and Pmin are maintained constant and attached to the same buttons.
During a session, the subject changes the button he presses in search of a variant that
would bring about maximum success. Figure 2. presents such a sequence of changes from
one button to the other during a session.


4. Conclusions

In spite of achievement in the field of decisions pertaining to systems characterised
by crisis, dedicated works does not comprise indications regarding attempts to broaden
the results or to apply them to the field of computer-aided education. Main contributions
and results in the field of computer-aided testing of maritime students are the following:
• Establishing an objective criterion to assess the capacity to adapt decisions and a
normal and logical behaviour under crisis and complex conditions .
• Designing and implementing a simulator meant to computer-aided testing of
maritime students to assess the decision making capacity under crisis conditions.


REFERENCES


Journal Articles

SKINNER, B. F., The Science of Learning and the Art of Teaching, Harvard, “Educational Review”, 1964.
University of Bucharest and Ovidius University of Constanta


280
SKINNER, B. F., „Teaching Machines”, Science Review nr. 27 (1958).


Conference Proceedings

DUMITRESCU, C., „ Theoretical Elements of Designing E-Training Process”, Annual meeting of scientific
works with international attendance – Strategies XXI-2006, E-Learning and Eduactional Software, 13-
14 April 2006, Bucureşti, Publishing House of National Defence, “Carol I” University, ISBN: 973-7854-
35-7 (13), 978-973-7854-35-3.
DUMITRESCU, C., „Requirements and Steps în Achieving Computer-Aided Education”, Proceedings of the
3
rd
Balkan Region Conference on Engineering Education, „Advancing Engineering Education”, 12-15
September 2005, Sibiu, Romania, ISBN: 973-739-147-0, http://brcee2005.ulbsibiu.ro/



The 3
rd
International Conference on Virtual Learning, ICVL 2008

281

Contents

MODELS & METHODOLOGIES


No Paper and Authors Page
1
Development of Group Division Algorithm And Discussion
Support System for Intra-class Discussions

Ikuo Kitagaki

Research Institute for Higher Education, Hiroshima University
2-12-1 Kagamiyama, Higashi-hiroshima, 739-8512, Japan
E-mail: kitagaki@hiroshima-u.ac.jp
101
2
THE VIRTUAL TRAINING CENTRE (VTC) FOR CNC
(COMPUTER NUMERICAL CONTROL)

Catalin Dumitras
1
, Sahin Mehmet
2
, Mihai Aura
1
, Yaldiz Süleyman
2
,
Bilalis Nikolaos
3
, Maravelakis Emmanuel
4


(1) Gh Asachi Technical University of Iasi, Faculty of Leather and
Textile Engineering, Romania
(2) Technical Science College, Selçuk University 42031, Konya, Turkey
mesahin@selcuk.edu.tr
(3) Department of Production Engineering & Management, Technical
University of Crete, 73100, Chania, Greece
(4) Design & Manufacturing Laboratory, Technological Educational Institute
of Crete, 73133 Chania, Greece
109
3
THE VIRTUAL TRAINING CENTRE FOR SHOE DESIGN
(VTC-SHOE): A MULTILATERAL VIRTUAL TRAINING
MODEL BASED ON A COMMON CURRICULUM

Aura Mihai
1
, Mehmet Sahin
2
, Süleyman Yaldiz
2
, Alina Dragomir
1
,
Nikolaos Bilalis
3
, Emmanuel Maravelakis
4

(1) Gh Asachi Technical University of Iasi, Faculty of Leather and
Textile Engineering, Romania
(2) Technical Science College, Selçuk University 42031, Konya, Turkey
mesahin@selcuk.edu.tr
(3) Department of Production Engineering & Management, Technical
University of Crete, 73100, Chania, Greece
(4) Design & Manufacturing Laboratory, Technological Educational Institute
of Crete, 73133 Chania, Greece
117
University of Bucharest and Ovidius University of Constanta


282
4
Virtual Learning Environments and World Languages
The Way Forward
The Flexi-Pack Project as SOAS-UCL CETL
(University of London)

Nathalie Ticheler

SOAS-UCL Centre for Excellence in Teaching and Learning Languages of the
Wider World, School of Oriental and African Studies Room 443
Thornaugh Square, London WC1H 0XG, UNITED KINGDOM
E-mail: nvticheler@yahoo.it
129
5
Applying 0/1 Integer Programming to Optimize User
Curriculum in a Virtual Learning Environment based on
Utility Function

Hamed Fazlollahtabar
1
*, Morteza Mofidi
2

(1) Young Researchers Club, Islamic Azad University Babol Branch, Iran
(2) Department of Industrial Engineering
Mazandaran University of Science and Technology, Babol, Iran
*E-mail: hamed@http.ac.ir
135
6
Selection of Optimum Maintenance Strategies in a Virtual
Learning Environment based on Analytic Hierarchy Process

Hamed Fazlollahtabar
1
*, Narges Yousefpoor
2

(1) Young Researchers Club, Islamic Azad University Babol Branch, Iran
(2) Department of Industrial Engineering
Mazandaran University of Science and Technology, Babol, Iran
*E-mail: hamed@http.ac.ir
143
7
Applying QFD Approach to Design an Online Course
in a Virtual Learning Environment

Hamed Fazlollahtabar

Young Researchers Club, Islamic Azad University Babol Branch, Iran
e-mail: hamed@ustmb.ac.ir
153
8
Multi-Criteria Decision Model for e-learning Architecture
Selection based on Utility Function and ELECTRE Method

Hamed Fazlollahtabar

Young Researchers Club, Islamic Azad University Babol Branch, Iran
E-mail: hamed@ustmb.ac.ir
161
The 3
rd
International Conference on Virtual Learning, ICVL 2008

283
9
Applying Integrated Strategic Planning and RADAR
Technique to Find Optimal Course Delivery Policy in a
Virtual Learning System

Hamed Fazlollahtabar
1
*, Ali Abbasi
2

(1) Young Researchers Club, Islamic Azad University Babol Branch, Iran
(2) Department of Industrial Engineering, Mazandaran University of Science
and Technology, Babol, Iran
*e-mail: hamed@ustmb.ac.ir
169
10
“Evalution traps”: a brief vademecum to avoid the most
common mistakes in distance learning evaluation

Gaetano Bruno Ronsivalle
1
, Piera Vivolo
2


(1) Professor of University of Rome, La Sapienza R&D Manager
AbiFormazione, Milan-Rome Italy R&D Manager LabelFormazione, Rome Italy
E-mail: sprsricercaesviluppo@abiformazione.it
(2) Researcher, LabelFormazione, Rome Italy 80, Scorticabove st, Rome,
IT – 00156 – ITALY
E-mail: pvivolo@abiformazione.it
177
11
Specifications of the "informatisation" processes
for productive-instructive workflows

Ioan Rosca
1
, Val Rosca
2


(1) LICEF institute- Teleuniversity of Montreal
E-mail: ioan.rosca@licef.teluq.uqam.ca
(2) Amazon development center – Iaşi
E-mail: rosca@amazon.com
185
12
Steps of Implementing an E-learning
Programme in Superior Education

Gabriela Moise
1
, Loredana Netedu
1
, Liviu IoniŃă
1


(1) Petroleum-Gas University of Ploieşti, no. 39 Blvd. Bucureşti, Ploieşti,
ROMÂNIA
E-mail: gmoise@upg-ploiesti.ro
199
13
Q-learning Approach in the Context of
Virtual Learning Environment

Ionita Liviu
1
, Tudor Irina
1


(1) “Petroleum-Gas” University of Ploieşti, 39 Bucharest Bd., 100680, ROMÂNIA
E-mail: iliviu@upg-ploiesti.ro
209
University of Bucharest and Ovidius University of Constanta


284
14
Analyzing Information Security Issues
Using Data Mining Techniques

Daniela Şchiopu
1
, Irina Tudor
1


(1) “Petroleum-Gas” University of Ploieşti, 39 Bucharest Bd., 100680, România
E-mail: daniela_schiopu@yahoo.com
215
15
Involving Learner’s Emotional Behaviors
in Learning Process As a Temporary Learner Model

Ahmad Kardan
1
, Younes Einavypour
1


(1) Advanced e-Learning Technologies Group, Department of Computer Engineering
and Information Technology, Amirkabir University of Technology (Tehran
Polytechnic), 424, Hafez St., Tehran, 15875-4413, Iran
E-mail: {aakardan, younos}@aut.ac.ir
223
16
Classification Based on Learner’s Ability and Emotionality
For Selecting a Suitable Teaching Method

Ahmad Kardan
1
, Younes Einavypour
1


(1) Advanced e-Learning Technologies Group, Department of Computer Engineering
and Information Technology, Amirkabir University of Technology (Tehran
Polytechnic), 424, Hafez St., Tehran, 15875-4413, Iran
E-mail: {aakardan, younos}@aut.ac.ir
231
17
Learning Object Tendency:
A New Concept for Adaptive Learning Improvement

Ahmad Kardan
1
, Samad Kardan
1


(1) Advanced e-Learning Technologies Group, Computer Engineering and
Information Technology Department, Amirkabir University of Technology (Tehran
Polytechnic), 424, Hafez St., Tehran, 15875-4413, IRAN
E-mail: {aakardan, skardan}@aut.ac.ir
237
18
Communication Models Used in the
Online Learning Environment

Gabriela Moise
1


(1) Petroleum-Gas University of Ploieşti, no. 39 Blvd. Bucureşti, Ploieşti, ROMÂNIA
E-mail: gmoise@upg-ploiesti.ro
247
19
Differential Geometry of Space Curves with Mathcad

Nicolae DăneŃ

Technical University of Civil Engineering of Bucharest
124, Lacul Tei Blvd., Bucharest, RO-020396, ROMÂNIA
E-mail: ndanet@cfdp.utcb.ro
255
The 3
rd
International Conference on Virtual Learning, ICVL 2008

285
20
Safety in Web 2.0

Dumitrescu Marina
1
, Dumitrescu Bogdan
2
, Daniel RăduŃă
3


(1) Technical College of Telecommunication „Nicolae Vasilescu-Karpen”,
Bacău, mimozad2005@yahoo.com
(2) Dumitrescu Bogdan – University of Bucharest, bogdumdum@yahoo.com
(3) Daniel RăduŃă – Security Department of Bitdefender AntiSpam,
Softwin – Bucharest, draduta@yahoo.com
265
21
Experimental Testing of Decisional

Carmen Dumitrescu
1


(1) Departament of Naval Electrotechnics, Electronics and Computers ScienceMaritime
University of ConstanŃa, 104, Mircea Cel Bătrân Street, ConstanŃa, România
E-mail: cdumit2002@yahoo.com
275



The 3
rd
International Conference on Virtual Learning, ICVL 2008

287



Sections

TECHNOLOGIES &
SOFTWARE SOLUTIONS


Technologies (TECH):
• Innovative Web-based Teaching and Learning Technologies
• Advanced Distributed Learning (ADL) technologies
• Web, Virtual Reality/AR and mixed technologies
• Web-based Education (WBE), Web-based Training (WBT)
• New technologies for e-Learning, e-Training and e-Skills
• Educational Technology, Web-Lecturing Technology
• Mobile E-Learning, Communication Technology Applications
• Computer Graphics and Computational Geometry
• Intelligent Virtual Environment

Software Solutions (SOFT):
• New software environments for education & training
• Software and management for education
• Virtual Reality Applications in Web-based Education
• Computer Graphics, Web, VR/AR and mixed-based applications for
education & training, business, medicine, industry and other
sciences
• Multi-agent Technology Applications in WBE and WBT
• Streaming Multimedia Applications in Learning
• Scientific Web-based Laboratories and Virtual Labs
• Software Computing in Virtual Reality and Artificial Intelligence
• Avatars and Intelligent Agents



The 3
rd
International Conference on Virtual Learning, ICVL 2008

289
Learning Distributed Activities Inside 3D Virtual Spaces

Dorin Mircea Popovici
(1,2)
, Jean-Pierre Gerval
(3)
, Felix Hamza-Lup
(4)
,
Norina Popovici
(1)
, Mihai Polceanu
(1)
, Remus Zagan
(1)


(1) OVIDIUS University of Constanta
124, Mamaia Bd, 900527, Constanta, ROMANIA
E-mail: dmpopovici@univ-ovidius.ro, norinapopovici@yahoo.fr, mythalic@gmail.com,
rzagan@univ-ovidius.ro
(2) European Virtual Reality Center, Brest, France
E-mail: popovici@enib.fr
(3) Institut Superieur de l’Electronique et du Numerique – Brest

20 rue Cuirassé Bretagne –

CS 42807 – 29228 BREST cedex 2 – FRANCE
E-mail: jean-pierre.gerval@isen.fr
(4) Computer Science, Armstrong Atlantic State University
Savannah, GA 31419, USA
E-mail: Felix.Hamza-Lup@armstrong.edu


Abstract
We discuss the effects of using distributed virtual reality environments in the
educational process. Pedagogical, technical and implementation aspects are
explored in conjunction with a virtual environment used in the engineering training
curricula.

Keywords: Virtual Reality, Learning, Teaching, Motivation.


Knowledge exchange occurs in the process of social interaction. The complexity of
the actual knowledge we are trying to assimilate makes us active parts in an educational
environment. Searching, discovering and testing, are the most frequent human activities
in such environments. When an abstraction level of knowledge is reached, these activities
are completed and creational acts may appear as complement to learning, with a
constructive feedback as one of the side-effects.
In the following sections we will emphasize the contribution of distributed virtual
environments to the learning process in order to transform it into an effective and
evolutionary one.


1. Introduction

Many educational virtual environments are using different metaphors in order to
facilitate the trainee learning on an abstract (math, physics, etc.) or concrete level (as
gesture or behavior in certain situations) but few of them are taking into consideration the
trainee motivation to ”do it”. Multimodal environments combining haptic feedback with
University of Bucharest and Ovidius University of Constanta


290
3D visualization (Hamza-Lup, 2008) prove very efficient at learning and understanding
concepts. Because the environments we propose are addressing mature users, motivation
may come not necessarily from the environment itself but from the user’s desire to
succeed in its integration within a professional context, very often being based on the
teamwork concept.
As students gradually gain confidence in the team they participate in, they become
autonomous and are willing to learn and acquire new knowledge; thus, they change from
being dependent to being independent and the relationships between individuals becomes
dynamic and warm. In particularly, this kind of environments is suitable for an
interdisciplinary team. For example, the system we discuss here, EngView, was
developed by a mixed team of computer scientists, engineers, and managers, as well as a
group of enthusiastic students. In this way, we have attained our main pedagogical
objective; that is to assure our students a rapid and successful integration in the
economical context. However, some difficulty will arise due to several factors. One of the
most important is the difference level of knowledge that the students attains during their
studies. Another is the student level of interest in the presented information and, a factor
with the same importance, the student motivation to learn.
There are different learning “speeds” and they vary from person to person. Often,
theory is easier to grasp than to translate into practice. Or vice-versa, practical skills
are quickly achieved, even without any basic understanding of the theory. In
spite of these difficulties, we want to achieve good theoretical and practical skills
employing such environments.
For theoretical knowledge level, the widely used method of multiple choice
examinations can be computer graded or easily marked with a template. However this
method does not provide any insight into the trainee’s work methods and adaptability. A
much better choice is the written examination. On the other hand, practical examinations
are somewhat more probing, however the trend is to have the candidate demonstrate his
skills on a simple application which can be easily and uniformly graded and which may
or may not be relevant to his industry (Roloson, 2004).
Because new technologies, as virtual reality (VR), facilitate learning through the
concepts construction based on the intuitions that arise from user direct experience of the
virtual environment (Bruner, 1986), we have decided to complete our teaching/learning
process by using these technologies. We are not eliminating multiple choice
examinations, but we consider that the communication and interaction within a
collaborative virtual environment may represent essential motivational dimensions of
trainee experience. This makes interaction, communication and motivation to be the most
important requirements of our technology.
Another important aspect is the reduced accessibility of the real setups for a group
of trainees. By switching the training sessions in real environment with training sessions
using virtual replicas, the trainee is able to obtain the confirmation of its practical results
obtained in virtual environment. So we are not eliminating the real tests in real situations,
but we let the students to exercise longer within a virtual setup, without any physical risks
and at lower costs for them (and for the setup of course). When they reach a certain level
of “virtual expertise”, they are allowed to prove this expertise in real environment.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

291
2. Success story

In order to demonstrate the effectiveness of such an educational virtual
environment we have implemented a training environment, called EngView (CERVA,
2007) that is basically a supplementary tool in the engineering curricula training process
in the domain of the non-destructive testing (NDT).


2.1. NDT principle

The most used formats in the NDT training process are A-scan, B-scan and C-scan
presentations, providing different ways of visualization and evaluation of the material
region being inspected. For our purposes, we have chosen to visualize only the C-scan method.
The high frequency ultrasonic C-scan presentation provides a plan-type view, depth
location, and size of the defects inside the probe; this makes C-scan a valuable tool to
monitor the precise location of defects between certain layers (see figure 1). The plane of
the image is parallel to the scan pattern of the transducer. C-scan presentations are
produced with an automated data acquisition system, such as a computer controlled
immersion scanning system.
The C-scan method is based on the transmission of a very high frequency signal (up
to 50 MHz) directed to the sample by a transducer. The sample and the transducer are
submerged in a coupling medium (water in our case). The initial signal is partially
reflected back to the transducer at interfaces, defects, porosities and at strong differences
in acoustic impedance in the sample and the rest of the signal. If not fully reflected, the
signal continues through the sample. In other words, between the initial pulse and the
back wall peaks there will be an additional peak caused by the sound wave going from
the water to the test material. This additional peak is called the front wall peak. The
ultrasonic tester can be adjusted to ignore the initial pulse peak, so the first peak it will
show will be the front wall peak. Some energy is lost when the waves hit the test material,
so the front wall peak is slightly lower than the peak of the initial pulse. In return, the
peak amplitudes as well as the time-of-flight of each returning signal are stored in a
computer data file and processed off-line to produce maps of the scanned area for the
sample placed at a particular depth.









Figure 1. C-scan principle

Scanning
Probe
Transducer
Increment
I
n
c
r
e
m
e
n
t

Scanning
University of Bucharest and Ovidius University of Constanta


292
2.2. Cooperation in a virtual environment

A predominant metaphor is the “virtual classroom”. In such a classroom, we
consider that organizing the learners in teams may reduce most of the gaps between the
individual knowledge by increasing communication and competition (in this order). In
this way competition becomes cooperation and increases the level of motivation. Hence,
the complexity that may arise in “natural” ways even in the most "simple" subjects is a
non-declared motivational factor when introduced gradually. Doing so, in the learning
context, if the learners’ needs are satisfied and their expectations are met, they will strive
to develop their professional competences. Indirectly they are contributing to the
development of the learning context (see figure 2).



Figure 2. Sample of a shared virtual classroom

Because the students are sharing the real environment, we want them to share the
same virtual environment also. Students naturally start to form small work teams in the
virtual setup, bases on the real environment configuration. Later, these teams will evolve
and will be based on the complementary knowledge that the team members posses, in
order to assure high-level of team performance. Shared experiences provide different,
perhaps even complementary perspectives to the lesson’s subject, depending on each individual.
The EngView setup was used during the second semester of 2007 in training
sessions by Engineering and Physics senior students, organized in 8 groups containing
25 students each. The NDT curricula require one practical evaluation on the basis of
6 laboratory hours. As mentioned before, the NDT makes no exception in both theoretical
and practical evaluation. To this end, the virtual environment contains pedagogical
resources that provide user access to theoretical background and evaluation as well as to
practical sessions.
The theoretical exam is organized on the basis of a multiple choice test containing
10 pure theoretical and 7 practical questions (CERVA, 2007). The students have 30 minutes
to answer all questions. The practical evaluation has three steps: the experiment
setup/calibration; the experiment itself; and the results interpretation. In the real
configuration, about 30 minutes are necessary for an experiment per student, without any
The 3
rd
International Conference on Virtual Learning, ICVL 2008

293
error recovery, so there is no possibility to try it twice. In this situation, it is often enough
that, during the exam, the student manipulate the real NDT setup, for the first time.
Because of the time limitation and the multi-user accessibility that a real configuration
presents, we decided to implement our own 3D-immersive simulation software, EngView.
By implementing all the functionalities of the real NDT installation, the EngView allows
the user to practice within an interactive 3D simulation of the real environment.



Figure 3. EngView – theoretical session (grila)



Figure 4. EngView – practical session

More precisely, the user is able to freely change the viewpoint inside the simulated
environment (front/back, left/right, up/down), as well as s/he may chose to navigate
through the simulated environment attached to the scanning virtual devices. This feature
University of Bucharest and Ovidius University of Constanta


294
allows the user to visualize the surface of the virtual scanned object during the simulation.
S/he can also move the three crane-like components of the virtual scanning device, in
order to virtually scan the simulated 3D object.
Students can reproduce different types of a real experiment using the EngView
system (preparing the sample, installing the type of transducer, setting the parameters of
moving engines to establish the type of scanning procedure). In order to realize a
comparative study, the user may choose even a particular moment in time. This way, the
students that work on the clients PC in EngView system are able to make the same kind
of analysis as in a real system.
The EngView system could be used independently, not coupled to the real system,
by installing it on a PC. This is an advantage since it allows students to train individually
using their home computer and an Internet connection.
We can observe (Figure 5) that keeping only the traditional training sessions (in
blue) is both non motivating and time-consuming. The possibility of failure because poor
practical skills and/or because of errors that may appear during the experiment is too high
for the current curricula. On the other hand, offering students the possibility of practicing
in a virtual configuration before the real one (in violet), we have succeed in motivating
them to prepare themselves. The students became more confident in their potential due to
the possibility to recover from errors and to experiment more training situations using the
virtual setup.

















Figure 5. Students training time completion in real configuration without virtual
training sessions (■), and with virtual training sessions (■)

In addition, the amount of training hours and practice established by regulations
could be considerably reduced. It makes seminars less expensive by using complex
immersive and interactive simulation with on-line tests. Moreover, it brings students
Students real training session time completion
0
20
40
60
80
100
120
Time [min]
N
u
m
b
e
r

o
f

s
t
u
d
e
n
t
s
Students training
classical mode
4 33 52 96 15
Students training
using virtual
equipment
64 81 32 23 0
10 - 15
min
15 - 20
min
20 - 25
min
25 - 30
min
more 30
min
The 3
rd
International Conference on Virtual Learning, ICVL 2008

295
closer to the practical part of their education and helps them acquire complete
comprehension of each method. This way, students will be able to do the job as soon as
they start working as employees.


3. Implementation issues

Our educational virtual environments are currently based on the assumption that
knowledge or skills acquired in a VR-based experienced environment will be transferred
to the real world. The effectiveness of such an experience depends on the user’s
capability to apply knowledge or skills acquired in both theoretical and practical mixed
experience VR to its real world counterpart.
The current learning materials are implemented using Moodle platform for the text and
multimedia resources (Word, PDF, PPT, AVI or JPEG files) as well as 3D virtual environments.
Concerning text-based and multimedia supports, we are exploiting the Moodle
(Moodle) facilities in order to align the pedagogical context to Sharable Content Object
Reference Model norms (SCORM). Moreover, we manage the users access to the
corresponding course materials, according to their curricula, and also the course
materials. This means that the tutors have the functional possibilities to create, modify
and publish educational materials (courses, seminars, home works, project subjects, tests,
others). Moreover, using such a system, the administrator is able to manage the courses,
the users, the groups of students, as well as the students enrolling in a specific courses.
The 3D environments are developed basically based on VRML (VRML) and/or
ARéVi API (Reignier et.al., 1998). ARéVi API has the advantage of being an open
source; it is C++ and OpenGL based, and is adaptive to very different configurations
starting from desktop systems, and ending with 3D stereoscopic immersion systems. In
order to put all together, we are using a reactive agent-based architecture (Popovici,
2004). This architecture assure the user immersion and evolution within the virtual space.
In order to ensure the distributed activities we have adopted the LAMP architecture;
i.e. Linux, Apache (Apache), MySQL (MySQL) and PHP (PHP) solution. Because our
educational environment is mostly 3D-oriented, we chose to create its architecture based
on the Ajax technology and X3D/VRML language. Ajax provides optimum update speed
between the client and the server by simulating a direct connection, while X3D has the
advantage of having an accessible structure that can be controlled from within the
Javascript (X3D). PHP and MySQL make the backbone of the whole application, being
the parts that are in charge of the user account and database management.


4. Conclusions

Many educational virtual environments are using different metaphors in order to
facilitate the trainee learning on an abstract (math, physics, etc.) or concrete level (as
gesture or behavior in certain situations) but very few of them are taking into
consideration the trainee motivation to ”do it”.
University of Bucharest and Ovidius University of Constanta


296
Within a group of children, each child contributes with its’ knowledge to the others
children’s knowledge, as (Johnson, 1999) suggested, cooperation between children in
virtual environments may have a positive effect on learning.
Because narrative lends itself to active exploration of a domain through challenging
and enjoyable problem-solving activities, (which are essential for learning) the children
develop communication and cooperation skills by invoking a narrative or game-like
context (Wood, 1996). Moreover, Internet, multimedia, VR and augmented reality (AR)
technologies are now part of our everyday life. These technologies facilitate learning
through the concepts construction based on the intuitions that arise from our direct
experience of the VE (Bruner, 1986).
We consider that the communication and interaction within collaborative multi-cultural
VEs are the most important motivational dimensions of user experience.


5. Acknowledgements

This work was partially funded by the INTUITION (FP6-IST-NMP-1-507248-2)
and EMULACTION (Fonds Francophones des Inforoutes – ref. no. 14G023) projects.
We would like to thank Mihai Iulian, Curcan Stefan and Gabriel Prodan for their
contribution to the virtual environment modeling and system implementation.



REFERENCES


Apache, http://www.apache.org
BRUNER, J. (1986), Actual Minds, Possible Worlds, Harvard University Press, Cambridge, MA.
CERVA, http://www.univ-ovidius.ro/cerva/engview , 2007.
HAMZA-LUP, F. G., SOPIN, I. (2008), "Haptics and Extensible 3D in Web-Based Environments
for e-Learning and Simulation", 4
th
International Conference on Web Information Systems
and Technologies (WEBIST), May 4-7, Funchal, Madeira, Portugal.
MySQL, http://www.mysql.com
Moodle, http://moodle.org
PHP, http://www.php.net
POPOVICI, D. M. (2004), Modeling the Space in Virtual Universes, PhD Thesis, Politehnica
University of Bucharest.
REIGNIER, P., HARROUET, F., MORVAN, S., TISSEAU, J., DUVAL, T. (1998), ARéVi: A
Virtual Reality Multiagent Platform, Lectures Notes in Computer Science, Volume 1434,
ISSN: 0302-9743 (http://www.cerv.fr/fr/activites/AReVi.php).
ROLOSON, C., ZIRNHELT, J. (2004), Performance Based Qualification: an NDT Service Industry
Perspective, paper no. 744, CD-ROM proceedings of the 16th WCNDT 2004 – World Conference
on NDT, Aug. 30-Sep. 3, Montreal, Canada, http://www.ndt.net/abstract/wcndt2004/744.htm
SCORM, Sharable Content Object Reference Model, http://www.adlnet.gov/scorm/
VRML, Virtual Reality Modeling Language, http://www.web3d.org
X3D, http://www.web3d.org

The 3
rd
International Conference on Virtual Learning, ICVL 2008

297
SVG Language (Scalable Vector Graphics)
For 2D Graphics in XML and Applications

Marin Vlada

University of Bucharest, 14 Academiei Street, RO-010014, Romania
E-mail: vlada@fmi.unibuc.ro, Web:www.ad-astra.ro/marinvlada

Abstract
The SVG technology is an open source copyrighted material of the W3C consortium
and it is a language for 2D graphics within the XML (eXtensible Markup
Language). The combination between SVG and JavaScript offers a powerful
platform usable for interactive 2D graphics, comparable to the Flash and Java
technologies. SVG offers XML graphics for the Web using three types of graphical
objects: vector graphic shapes (lines and curves), images and text. Objects may be
grouped, transformed and represented, dynamically and interactively. The SVG
uses XML text standards, JPEG and PNG image formats, DOM (Document Object
Model) for scripting and interactivity, SMIL for animation and CSS for styling. The
present paper constitutes a presentation of the SVG and it also describes a few
applications written in SVG.

Keywords: SVG Technology, 2D Graphics, XML, Document Object Model.


1. Introduction – Web 1.0 and Web 2.0

MOTTO: "Things to watch: SVG – Scalable Vector Graphics – at last, graphics
which can be rendered optimally on all sizes of device" (TIM BERNERS-LEE,
inventor of the World Wide Web www.w3.org/People/Berners-Lee)

The complexity of computer applications in various fields (including education),
has powered the improvement of both operation systems and programming languages, as
well as the improvement of technologies and platforms. New operation systems, new
programming languages and new technologies have been conceived and developed. If the
invention and putting into use of the microprocessor in the ‘70’s meant a true revolution
in the field of computer hardware architecture, the 90’s marked a true revolution in the
field of both computer networks as well as the field of programming languages (through
the advent of Java and JavaScript) and operating systems (Linux, Windows). Thus,
Web technologies appeared. One should also mention the development and evolution of
the C++ language which during the 80’s implemented and developed the object oriented
model of programming (the objectual programming model is rooted in the SmallTalk or
Lisp as well as other programming languages) and also object oriented programming
(OOP – Object Oriented Programming) [8].
University of Bucharest and Ovidius University of Constanta


298
At the beginning of the 90’s, the HTML (Hypertext Markup Language) appeared a
thing that determined the dissemination of static Web pages as well as an explosive
development of the WWW (World Wide Web) system. The need to develop dynamic Web
pages has determined the advent of various other technologies such as: JavaScript,
JavaServer Pages (JSP), VBScript, PHP, ASP, Macromedia Dreamweaver, a.s.o.,
which were mainly meant for server side applications, while others were meant for client
side applications. In the field of dynamic and interactive graphic applications the last 10
years were dominated by the Java and Flash technologies: Web 1.0 generation.
As a consequence of the developments offered by the XML (eXtensible Markup
Language), JavaScript language, DOM (Document Object Model) for scripting and
interactivity, SMIL (Synchronized Multimedia Integration Language) for animation and
CSS for styling, in 2003, the W3C (Worldwide Web Consortium – W3C) elaborates the
1.1 SVG (Scalable Vector Graphics) specification – http://www.w3.org/Graphics/SVG/.
This is an open source platform and is a consistent alternative to the Java and
Flash technologies. „SVG is used in many business areas including Web graphics,
animation, user interfaces, graphics interchange, print and hardcopy output, mobile
applications and high-quality design” [7]. The authors of this platform are: Adobe, Agfa,
Apple, Canon, Corel, Ericsson, HP, IBM, Kodak, Macromedia, Microsoft, Nokia, Sharp
and Sun Microsystems.
The examples and graphical applications shall be viewed (using a compatible
browser) with Firefox 1.5+, Opera 9 or Internet Explorer with Adobe SVG plug-in
(Adobe SVG Viewer [1]). Since 2003, appears the generation Web 2.0: new technologies
and Web services.
Using the SVG platform, in the field of education, a number of graphical
applications for various sciences (mathematics, physics, information technology,
chemistry, a.s.o) can be imagined and developed, so as to aid in the presentation of
various phenomena, terms and concepts. The presentations can be made using a computer
and a video projector installed in the classroom/ room where pupils/ students assist the
lesson. The dynamic aspect and interactivity of the presentation make these presentations
attractive. The teachers and IT specialists must cooperate, with the final purpose of
developing such graphical applications. The classic textbook for a certain subject is
unable to offer the level of interactivity and dynamics that are offered by the SVG,
therefore it is desirable that the textbook is completed through the use of adequate
software products (educational software), a software that will meet the requirements of
pupils, students and educators. While a classical textbook constitutes the work of one or
several specialized authors, a software product which is meant for educational use can
only be completed through collaborative work of several specialists from various fields:
education, IT, psychology, learning sciences and even pupils or students. Thus,
lessons and classes shall be at the same time attractive and useful, to the purpose of obtaining
the competencies that are taken into consideration through the adequate curricula.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

299
2. Scalable Vector Graphics (SVG)

SVG is a platform for describing the 2D graphical applications within the XML
language. The combination between the SVG and JavaScript offers a powerful platform
for interactive 2D graphics, comparable to the Flash and Java technologies. The SVG
platform offers XML graphics for the Web through three types of graphic objects: vector
graphic shapes (lines and curves), images and text. The objects can be grouped,
transformed and represented dynamically and interactively. SVG uses XML standards for
text, the JPEG and PNG formats for images, DOM (Document Object Model) for
scripting and interactivity, SMIL (Synchronized Multimedia Integration Language) for
animation and CSS for styling. The SVG platform consists of two parts: a basic XML
type file and API programming for the 2D graphic applications. „Key features include
shapes, text and embedded raster graphics, with many different painting styles. It
supports scripting through languages such as ECMAScript and has comprehensive
support for animation” [12]. An in-depth familiarization with the HTML, XML as well as
object oriented programming (OOP) will assist in a clearer understanding of the usage of
the SVG specs [14]. The XML language (eXtensible Markup Language) is the language
that offers a format for storing and transmitting data through a declarative description.
The XML „grammar” includes XHTML (the XML version of HTML), SVG, MathML
(the Mathematics Markup Language), ChemML (the Chemistry Markup Language) and
GML (the Geography Markup Language).
As a scripting language, through the use of JavaScript and SVG Document Object
Model (SVG-DOM), SVG is an extension of HTML DOM level 2, which is highly
familiar to all Web developers. SVG elements can be characterized by animation through
the use of Synchronized Multimedia Integration Language (SMIL).
Just like XHTNK and MathML, SVG is of XML type; all SVG files and documents
share a .svg extension and can be edited using a plain text editor, (such as for instance
Notepad). These files are not compiled; instead they are merely interpreted by the
browser (Firefox 1.5+, Opera 9 or Internet Explorer with Adobe SVG Viewer).
Document Type Definition (DTD) is the description of elements and attributes
(svg11.dtd) corresponding to the SVG type statements, which are subsequently used
in graphic applications ("http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd"). The
„root” elements of the XML documents are those XML tags that are interpreted by the
browser. For instance, <html> for XML and <svg> for SVG. „Namespace” XML or
"xmlns" attributes will perform a unique identification description of the SVG attributes
using XML data. SVG applications need the following statements:
• xmlns = http://www.w3.org/2000/svg
• xmlns:xlink = http://www.w3.org/1999/xlink


University of Bucharest and Ovidius University of Constanta


300

XML document header DTD SVG document header „Root “elements













SVG Content/ elements Drawing window (560 × 480 pixeli)

For SVG version 1.0 DTD reference (svg10.dtd) is:

<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN"
"http://www.w3.org/TR/2001/REC-SVG20010904/DTD/svg10.dtd">

Development of graphic applications using SVG implies a knowledge of SVG
specs and use of definitions and attributes in accordance with the XML and SVG
definitions. All SVG content shall be enclosed between the <svg> tags. It can be taught
through examples.


2.1. Programming in SVG

The essential elements (tags) for graphical apps are: <circle>, <ellipse>,
<rect>, <line>, <polyline>, <polygon> and <path>. An all important
tag is also <g> for grouping elements (shapes) and <use> employed in reusing of the
elements predefined in the <defs> section. A complete list regarding definition, syntax
and use of SVG elements can be found at the W3C SVG 1.1 Recommendation
(www.w3.org/TR/SVG11/) web address [13] and also at the W3Schools SVG site [15]. As
an example, we are herewith presenting the SVG code (ex1.svg) that performs
generation of geometric shapes: polygon, circle, rectangle, line (straight) and path, while
indicating for each of them the corresponding attributes and elements of identification.
<text> element is also to be employed for creating text.




<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">

<svg xmlns="http://www.w3.org/2000/svg"
xmlns:xlink="http://www.w3.org/1999/xlink"
width="560" height="480" >
...
Content svg
...
</svg>

The 3
rd
International Conference on Virtual Learning, ICVL 2008

301


Drawing of geometrical shapes and text

Ex1.svg (based on an example by David Lane [4])



















If it is desired that the text generated shall include a certain typeface of a specified
size, a class shall be defined through a CSS style sheet (class="title") (which
shall be introduced by the <polygon> tag) in the <defs> section:










In this case, the <text> element shall be:

<text class="title" x="270" y="50">Example 1</text>



<?xml version="1.0"?>
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">

<svg xmlns="http://www.w3.org/2000/svg"
xmlns:xlink="http://www.w3.org/1999/xlink"
width="570" height="470" >
<polygon style="stroke:#24a;stroke-width:1.5;fill:#eefefe"
points="10,10,400,10,430,230,10,280,10,10" />
<circle style="stroke:#d33;stroke-width:2;fill:#7ce"
cx="100" cy="80" r="50" />
<rect style="stroke:#2aa;stroke-width:7;fill:#ded;opacity:.8"
x="170" y="80" height="120" width="220" />
<line style="stroke:#eea;stroke-width:6" x1="30" y1="250"
x2="340" y2="60"/>
<path style="fill:#daa;fill-rule:evenodd;stroke:none"
d="M 230,250 C 360,30 10,255 110,140 z "/>
<circle style="stroke:#d33;stroke-width:2;fill:#7ce"
cx="370" cy="140" r="30" />
<text x="270" y="50">Example 1</text>
</svg>

<defs>
<style type="text/css"><![CDATA[
.titlu { font-size:30px; font-weight: bold; font-family:
batang;
stroke: none; fill: black; text-anchor: middle}
]]></style>
</defs>
University of Bucharest and Ovidius University of Constanta


302
Changing coordinates (changing to user defined coordinates)

In the preceding example, coordinates are expressed in pixels and their value is
relative to the origin (0,0) which in SVG (as well as in other languages) is implicitly the
upper left corner, while positive X coordinates are to the right as related to origin and the
Y coordinates are downwards [9, 13]. Certain calculations must be performed within the
applications, so that all coordinates that are used should be relative to this origin. In order
to avoid a large amount of mathematical calculations, homogenous coordinates (x,y,w),
are used, whereas (x,y) are Cartesian coordinates, and w is the multiplication factor. 2D
transformations are widely used in computer graphics (scaling, translation, rotation), by
representing them as a multitude of 3 – dimensional linear transformations.
Such representation is defined by a simple 3 x 3 matrix:



This way, each transformation has its own transformation matrix which is used in
calculations. The result is that all transformation is represented by a matrix calculation:
the (x, y, w) line vector is matriceal multiplied with the transformation matrix.
As an example, should we have to move the default origin (0, 0) to the point
described by coordinates (250, 250) and change the Y positive axis to the upwards
direction (as it is generally represented in mathematics), the following are needed: a = 1,
b = c = 0, d = – 1, e = f = 250, thus the transformation matrix is matrix(1, 0, 0, – 1, 250,
250). For further details: "Coordinate System Transformations" (Sec 7.4) -
http://www.w3.org/TR/SVG/coords.html [4,13]. In SVG this is achieved through the use
of the <g> element and the „transform” attribute:






Drag and Drop using SVG

SVG uses Document Object Model (http://www.w3.org/TR/SVG/svgdom.html)
through which it includes OOP JavaScript techniques with the purpose of generating
objects, properties and methods. More details regarding the description of these
properties and methods can be found in chapter 5 of Document Structure (http://www.w3.
org/TR/SVG/struct.html). JavaScript code is introduced using <script> tag,
subsequent to the opening of the <svg> tag, but preceding the <defs> section:


<g transform="matrix(1, 0, 0, -1, 250, 250)">
...
svg elements centered in (250,250)and the positive Y upwards
...
</g>

<script type="text/ecmascript">
//<![CDATA[
...
Javascript code
...
// ]]>
</script>

The 3
rd
International Conference on Virtual Learning, ICVL 2008

303








3. Graphic applications developed using SVG

Dynamic and interactive aspects and facilities offered by SVG offer a distinct
possibility of developing graphic dynamic and interactive applications for various items
of study. These applications can be used by the educators and trainers (teachers) within a
class of pupils/ students in order to explain various phenomena, concepts, terms through
direct involvement of the user in understanding theoretical and practical aspects of the
subject/ theme approached. This approach creates an environment which is similar to an
experiment which actually constitutes the very basis of the learning process. Learning
sciences and psychology studies demonstrate the fact that the pupil/ student is actively
involved in the activities that request performing experiments, analysis and
interpretations of phenomena, concepts, terms. Within the learning process, this is
deemed as an important step to learning through discovery.
Object Oriented Programming (OOP), including events generated through the use
and programming of mouse movements, animation programming using SMIL
(Synchronized Multimedia Integration Language), DOM (Document Object Model), use
of Java technology and „Drag and Drop” technique are some of the essential parameters
that define the advantages of the SVG platform. We recommend the study of the source
code of applications developed by David Lane [4]: Thales.svg (triangle inscribed in a
half-circle; modification of the triangle apex position on the circumference using the
mouse); UnitCerc.svg (the unit circle represented in either ortho or polar coordinates;
indicating a point on the circumference using the mouse, calculating the radius angle in
degrees, rads and also values of the cos and sin functions); ParamPlot.svg (drawing a
curve using parametric equations – generating trajectory is indicated as well as the
manner of drawing the curve). Below we present various applications by the parametric
equations of curves.

University of Bucharest and Ovidius University of Constanta


304





Pascal's snail (it is proposed that the experiment / demo):
function xFunc(t){
return 2*(a*Math.cos(t)+ b)*Math.cos(t);}
function yFunc(t){
return 2*(a*Math.cos(t)+ b)*Math.sin(t);}
...
function makePlot(){//function of drawing a curve by meshing [4]
var dstring="M"+xFunc(0)+","+yFunc(0)+" ";
for(t=0;t<nplotpoints;t++){
dstring=dstring+"L"+xFunc(period*(t+1)/nplotpoints)+","+yFunc(peri
od*(t+1)/nplotpoints)+" ";
}
fPlot.setAttributeNS(null, "d", dstring);

function xFunc(t){
return (a+b)*Math.cos((b/a)*t) -
b*Math.cos(t+(t*b)/a);
}
function yFunc(t){
return (a+b)*Math.sin((b/a)*t) -
b*Math.sin(t+(t*b)/a);
}

Epicicloida
function xFunc(t){
return a*t*Math.cos(t);
}
function yFunc(t){
return a*t*Math.sin(t);
}
Spiral of Archimedes
The 3
rd
International Conference on Virtual Learning, ICVL 2008

305
NOTE. Drawing a curve (Whether C is a curve that borders the field D. The curve
C is modeled using the polygon line P = P
1
…P
n
, P
i
(x
i
, y
i
), i = 1,n), C = FrD = Imγ ,
γ class route C
1
upon portions,
R
b a
2
] , [ : → γ , )) ( ), ( ( ) ( t y t x t
i
= γ , a ≤ t ≤ b.
Using the bijective application between the real segments [0,1] and [a,b], given by
φ(t) = a + t(b – a), the polygon line is modeled using the reunion of the γ
i
routes
parametrically represented as follows:

2
] , [ : R b a
i
→ γ , )) ( ), ( ( ) ( t y t x t
i
= γ , i=1,n,

whereas

x(t) = x
i
+ t (x
i + 1
– x
i
), y(t) = y
i
+ t (y
i + 1
– y
i
), i = 1, n – 1

noting that for the last route, γ
n
parametric equations are

x(t) = x
n
+ t (x
1
– x
n
), y(t) = y
n
+ t (y
1
– y
n
).





4. Conclusions

a) "SVG is a language for describing two-dimensional graphics in XML. SVG
allows for three types of graphic objects: vector graphic shapes (e.g., paths consisting of
straight lines and curves), images and text. Graphical objects can be grouped, styled,
function xFunc(t){
return (a-b)*Math.cos((b/a)*t)
+ b*Math.cos(t-(t*b)/a);
}
function yFunc(t){
return (a-b)*Math.sin((b/a)*t)
- b*Math.sin(t-(t*b)/a);
}
Hipocicloida
University of Bucharest and Ovidius University of Constanta


306
transformed and composited into previously rendered objects. Text can be in any XML
namespace suitable to the application, which enhances searchability and accessibility of
the SVG graphics. The feature set includes nested transformations, clipping paths, alpha
masks, filter effects, template objects and extensibility." [7].
b) „Scalable Vector Graphics (SVG) is the open source Worldwide Web Consortium
(W3C) recommendation for two dimensional vector graphics. The combination of SVG
and JavaScript is a powerful platform for creating interactive graphics, comparable to
Flash and Java.” (David Lane) [4].



REFERENCES


[1] Adobe (2005), SVG Viewer, http://www.adobe.com/svg/viewer/install/main.html
[2] TIM BERNERS-LEE, inventor of the World Wide Web, Personal Web Page, www.w3.org/
People/Berners-Lee, accessed august 2008.
[3] DIANA DIACONU (2006). WebPages using JavaScript, EduSoft Publishing House.
[4] DAVID LANE (2007), Scalable Vector Graphics, Published February 2007; article ID
1381,accessed august 2007, http://mathdl.maa.org/mathDL/4/?pa=content&sa=viewDocument
&nodeId=1381,
[5] SVGOPEN Organization, http://www.svgopen.org/2007/aim_en.shtml
[6] SVG Foundation, http://www.svgi.org , accessed august 2008
[7] SVG Wiki, http://wiki.svg.org/Main_Page/, accessed august 2008
[8] MARIN VLADA (2006), From Green’s Theorem to Computational Geometry, CNIV-2006,
National Virtual Learning Conference, Educational Software, 4th Edition, 27-29 October
2006, Publishing House of the University of Bucharest, http://fmi.unibuc.ro/cniv/2006/,
accessed august 2008.
[9] MARIN VLADA, ADRIAN POSEA, IOAN NISTOR, CĂLIN CONSTANTINESCU (1992).
Computer Graphics Using Pascal and C Languages, vol. I, II, Tehnica Publishing House, Bucharest.
[10] Yahoo svg newsgroup, http://tech.groups.yahoo.com/group/svg-developers/
[11] JOHN C. WHELAN, KELLY CAREY (2005), SVG For Teaching 2D Graphics Standards,
http://www.svgopen.org/2005/papers/TeachingGraphicsStandards/index.html
[12] W3C, SVG – Scalable Vector Graphics, http://www.w3.org/Graphics/SVG/
[13] W3C, Scalable Vector Graphics 1.1 Specification, http://www.w3.org/TR/SVG/SMIL
[14] W3C, XHTML, MathML, http://www.w3.org/markup/, http://www.w3.org/math/
[15] W3Schools, SVG, http://www.w3schools.com/svg/default.asp, accessed august 2008.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

307
Interactive Informative Unit Based on Augmented Reality
technology

Dorin Mircea Popovici
(1,2)
, Mihai Polceanu
(1)


(1) OVIDIUS University of Constanta
124, Mamaia Bd, 900527, Constanta, ROMANIA
E-mail: dmpopovici@univ-ovidius.ro, mythalic@gmail.com
(2) European Virtual Reality Center, Brest, France
E-mail: popovici@enib.fr


Abstract
In the present contribution we shall introduce the IIUBAR system, which represents
an Interactive Informative Unit Based on Augmented Reality technology (AR),
provided by the Augmented Reality Toolkit (ARToolkit) API. After invoking the
potential of the proposed system in educational contexts, theoretical concept behind
the ARToolkit package as well as advantages and disadvantages of the possible
solutions will be brought to discussion.

Keywords: Augmented reality, interactive interface, educational virtual environment.


Learning and teaching are two complex, continuous and complement processes.
Started from the early ages, learning has to find continuously motivational factors in
order to become effective and evolutive. Only when an abstraction level of knowledge in
a specific domain is reached, then teaching may appear as complement to learning, with a
constructive feedback as one of the side-effects.


1. Introduction

The widest dissemination form of learning environments in our days still remains
the World Wide Web. It is most probably because this technology represents the basis for
the online education and it offers knowledge to students independently to the moment, the
place and the time duration of the learning act.
It doesn’t matter if we talk about static or dynamic Web-based pages or if the
learning environment is a multimodal (multimedia) 2D or 3D metaphor of a real pedagogical
situation that is too dangerous or too expensive to be re-created basis on real means.
What we think that really matters is the possibility to transform the usual actors,
students and teachers, into involved active actors. To this end, we want to install them the
sensation that they represent active parts of the learning/teaching process. This way we
catalyze the creational state-of-mind as well as the self-confidence at individual level as
premises of long-duration collaboration between the individuals.
University of Bucharest and Ovidius University of Constanta


308
Other human related aspects have to be taken into account, as motivation, emotion,
satisfaction, and error, at individual level as well as at competitive or collaborative levels.
In this matter, not rare are the situations in which the lack or inaccessibility of
information becomes a stress factor in everyday life. If for a person who possesses all
physical means this incommodity is easy to overcome, for a handicapped person this
could represent a real challenge.
The technology of the third millennium makes itself noticed more and more in the
design of assistance systems for people with disabilities.
Without any distinction between these two categories of users, in the current work
we present an original solution based on augmented reality technology: the implementation of
an interactive information unit (IIUBAR) which targets a wide range of people, usable
both in informative and educational situations.
In the following we present our reasons in using mixed realities in public
educational and informative contexts. Next, after a brief review of some of the most
important achievements in the field of AR, we offer a general description of the IIUBAR
system followed by a section destined to discussions. We close with conclusions and
possible directions of development of the IIUBAR system.


2. Potential impact on virtual learning

Many of the current solutions for virtual or augmented reality systems are designed
to be manoeuvred by a single user, thus limiting the access to the application. When using
such an application in virtual learning, the experience gained by the user will only be
from the sole interaction with the system. From general experience, learning efficiency
grows directly proportional with communication, collaboration and motivation.
To achieve these key factors of learning, one faces the problem of designing the
learning system. As the three factors are human in nature, the system must include more
than one user (learner) to achieve the desired efficiency.
Only few of these systems take a group-oriented approach regarding the learning
efficiency issue. Having a group-oriented system would ensure that an exchange of
information between users will occur, and that the individual experiences of one user may
complement the ones of another, and vice-versa; in this way, users practically participate
in the development of the learning environment, while gathering more information, and
keeping themselves highly motivated through this process. Paraphrasing, the users
become teachers for each other, and information flow is eased.
The IIUBAR system may be used by groups of users with the role of enhancing the
experience gained from the learning material. Among the possible uses of the system are
the following three scenarios (but not only).
Generally, the system may be used in any public spaces as interactive and
informative access point (see figure 1). In particular, the virtual information stand in
thematic museums may be used to animate the static nature of these locations, making
them more attractive to visitors, and simultaneously providing the user with extra
information (see figure 2). Of course, the potential of the IIUBAR may be proved also in
The 3
rd
International Conference on Virtual Learning, ICVL 2008

309
interactive classes by providing 360
o
panoramas with unlimited possibilities in the matter
of content. For example, the system can be used in geography classes to give the learners
an experience of some of the places they have never visited before; the impact on the user
would be greater than a regular image, because of the simple but effective physical action
of rotating the viewpoint in the virtual environment.



Figure 1. IIUBAR in practice: up-overview, down-detailed view



Figure 2. Using IIUBAR in museums – detailed view


3. Presentation of the technology used

One of the challenges that the development of the mixed reality environment poses
is determining the position of the observer. This information needs to be extracted from
the image taken with a camera. To resolve this issue, ARToolkit (ARToolkit, 2007) relies
on computer-vision algorithms with which it can detect and recognize markers (the
symbols are each associated with a 3D model inside the application). These algorithms
are similar to those used by the artificial neural network technology in that of processing
signals in general, and of images or sounds, in particular.


University of Bucharest and Ovidius University of Constanta


310
3.1. Theoretical basics – the functioning mechanism

In the following section we shall describe the steps taken in the algorithm that
ARToolkit uses for image processing:
• The first step consists of capturing the image.
• After which applying a filter is necessary for localizing the marker. This filter is
also known as the binarization step, in which the initial data is transformed, by colour
differencing, into a black and white picture that can now be easily represented with binary
code, and thus stored for further processing.
• The edge detection takes place: the contour lines’ positions are estimated, which
are afterwards parameterized.
• The marker’s corners are calculated at a sub-pixel level.
• The obtained image is normalized.
• The known symbols are loaded into the application and associated with the real situation.
• The marker’s homography is calculated (Homography is a mathematical concept
that is defined as the relation between two geometrical figures, so that any point in one of
the figures corresponds to one and only one point in the other, and vice-versa).
• Relative transformations between the camera and the marker are calculated, and
optimizations are applied. This results in parameters which are to be used by the
application.
This algorithm is repeated for every frame of the video capture.
Latest attempts to overcome the limits of the current solution imply the use of neural
networks to solve the problem of marker detection and interpretation (Gomez et.al, 2007).
The main strength of neural networks is correlating information with pre-made
templates and even adapting the templates to environment changes.
The learning and pattern recognition abilities are possible because of the adaptive
sensibility of the neurons that are the basic elements of these networks, sensibility also
known as adaptive threshold.
Using this technology may revolutionize the adaptability of the system, making it
resilient to luminosity changes that are so frequently encountered in real scenarios and
giving it the ability to recognize also real objects, apart from markers.


3.2. Current solutions

ARToolkit is a software library meant for creating augmented reality, based on the
2D and 3D graphics API named OpenGL. Usage of ARToolkit is in continuous growth at
international scale for the possibilities that it offers and the open-source license under
which it is distributed.
The prime strength of ARToolkit is to make the direct interaction between the user
and the virtual environment more intuitive. This way, occasional users who do not
possess technical skills may find the system accessible.
Among the applications built based on this platform we can mention (ARToolkit, 2007):
The 3
rd
International Conference on Virtual Learning, ICVL 2008

311
• ARCO: Augmented Representation of Cultural Objects – a research project
financed by the European Union having the goal of developing techniques for museums
to create 3D models of the exhibitions online (ARCO, 2003),
• Component assembly or machinery manipulation environments with applicability
in a wide range of domains such as physics, chemistry or constructions (FaiMR, 2003;
ARToolkit, 2007),
• Applications in virtual learning, interactive lessons which help pupils or students
to gain a better understanding of the phenomena that they study; an example in this sense
would be the virtual theatre meant for children of pre-scholar and scholar age, presented
in (Popovici et.al, 2006),
• Games and other entertainment or educational applications (Billinghurst et.al.,
2001; Park and Woo, 2005).
Nevertheless, ARToolkit API it is not the only available technology in augmented
reality. Just think the virtual reconstructions of historical sites (Vlahakis, 2002;
Papagiannakis, 2004).


4. About the IIUBAR system

IIUBAR exemplifies an easy-to-use informative and educational system. It is
characterized by a natural, real-time interaction with the user. The system is formed of a
rotating stand, on which a computer (laptop) is placed that processes the information
received from two webcams to create a mixed environment by superimposing a 3D
virtual model over the real images (see figures 1 and 2).
When implementing the IIUBAR system, there were three different solutions taken
into consideration.
• Mobile system with one marker per objective,
• Fixed system with one marker per objective,
• Fixed system with one marker per environment.
In the following paragraphs we consider useful the observations made based on the
analysis of the implementation possibilities of the system. Thus we will specify details
about each solution and will argument the solution chosen for final implementation.


4.1. Mobile system with one marker per objective

The system configuration in this case assumes that a marker is created for each and
every objective from the real world, and that a 3D model is associated with each marker.
The hardware solution relies on linking and synchronizing a webcam and a HMD (head
mounted display) with a laptop that would be bared by the user (see figure 3.a).
University of Bucharest and Ovidius University of Constanta


312


a) b)

Figure 3. a) Mobile; b) Fixed – configuration with one marker per objective

Advantages: Because of the portability of the HMD, this configuration proves a
high level of mobility, without binding the user to one point.
Disadvantages: As the HMD, and thus the camera, moves simultaneous with the
user, the stability of the captured images is below average. Because these images are the
only modality for the user to perceive the environment, this version is not feasible.


4.2. Fixed system with one marker per objective

The configuration in this case is simplified in the sense that a touring table is used
in place of the HMD (see figure 3.b).
Advantages: The lack of the trepidations that we found in the previous version
makes this configuration superior in the sense of image clarity.
Disadvantages: Having a fixed point from which the detection is made, the markers
found far away from the system are not clearly detected, nor able to be visualized. Thus
the system’s capacity to provide the solicited information decreases dramatically.


4.3. Fixed system with one marker per environment

To combine the advantages of the two previous solutions, for the final implementation
we opted for the fixed solution with a single marker per environment.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

313



Figure 4. Fixed configuration using a single marker

This version assumes two modifications: one at the hardware level and the other at
the software level. The hardware change is represented by an additional camera, so that
we are able to split assign each the roles of harvesting real images and the other of
detecting the rotation angle relative to the marker. The software change is executed
through three important steps:
• Adapting the standard ARToolkit application named “twoview” so that using
models created in VRML (Virtual Reality Modeling Language) format is possible.
• Changing the transformation matrix used for storing the position parameters of
the marker. This is necessary because the rotation of the 3D model has to be done around
a different axis than the one of the markers rotation.
• Information received from the detecting camera (webcam 2 from figure 4) must
be used to generate the 3D model over the images taken from the camera that observes
the real environment (webcam 1 from figure 4), unlike the example in which each camera
was used for a separate image interpretation process.
The present configuration benefits from a 360
o
vision field, by using the rotating
stand that was introduced in the previous version, and from an significantly superior
image clarity and resolution, made possible by combining the images of the two webcams.
The first webcam is placed to receive real images on a horizontal direction, while
the second is directed vertically to detect the marker that is placed above the IIUBAR
system (Figure 4).
Webcam 2 will supply the application with the rotation parameters of the system
relative to the marker. With the help of these parameters, the viewpoint of the user can be
determined. Thus the correct superimposing of the 3D model becomes possible, so that a
complete interactive environment is obtained that responds to the users actions in real-time.
Advantages: Because the way in which it was design, this system has a high
sensibility to the user’s orientation; it is adaptable to the environment in which it is placed
(changing the 3D model one can satisfy the requirements of any new environment); it
benefits from a very high image stability.
University of Bucharest and Ovidius University of Constanta


314
These qualities gain more importance when placing the system into a real
collaborative context, this is exactly the case of group visits in a real scenario (for
example an university campus).
Disadvantages: The lack of mobility because of the fixed stand on which the
system is placed, and the necessity of installing it indoors.


5. 3D environment

The language used for building the virtual 3D models is VRML (Virtual Reality
Modeling Language). There are several reasons for which we opted for this language.
Firstly, the VRML standard is supported by the ARToolkit platform and permits the rapid
description of complex geometries, unlike OpenGL. Secondly, using this file format, the
IIUBAR system is easily adaptable to different configurations of the real environments.
An exceptionally important factor in creating the model is relating it to the space in
which the system will be placed. For an optimum functionality, prior measurements are
required to determine the distances between the positions where virtual objects are to be
placed. The lower the measurement error is, the higher the precision with which the
virtual and real worlds overlap (see figure 5).


Figure 5. The virtual 3D model and the scenario schema
The 3
rd
International Conference on Virtual Learning, ICVL 2008

315
Also another aspect that must not be overlooked is the camera calibration. Without
a correct calibration, the 3D model may not entirely fit the environment, because the
optical distortion produced by the camera. This problem can be solved by using the
ARToolkit utilities named „calib_camera2”, „calib_param” or „calib_distorsion”.


6. Conclusion

We consider that the potential of this platform is well above average. Firstly, it
requires a very low level of knowledge for proper use. Secondly, the assembly costs are at
minimum. Thirdly, it proves a great level of accessibility and also the application may be
adapted to contain more information, for user satisfaction.
The system may be successfully used as a public information stand, inside
institutions. Although it has many advantages, the IIUBAR system’s drawback would be
the fact that it can only be installed indoors.


7. Acknowledgements

The present system has been implemented in the Laboratory of Virtual and Augmented
Reality Research (CERVA, 2007) and is supported by the TOMIS project, no: 11-041/2007,
by the National Centre of Programs Management, PNCDI-2 – Partnerships program.



REFERENCES


ARCO, http://www.arco-web.org/TextVersion/Description/Description1.html, 2003.
ARToolkit, http://www.hitl.washington.edu/artoolkit/, 2007.
BILLINGHURST, M., KATO, H., and POUPYREV, I., The Magicbook – Moving Seamlessly Between
Reality and Virtuality, IEEE Computer Graphics and Applications, (May-June): 2-4, 2001.
CERVA, http://www.univ-ovidius.ro/cerva, 2007.
FaiMR – Furniture Assembly Instructor in Mixed Reality, http://staff.fh-hagenberg.at//haller/research-
faimr.html, 2003.
W. L. GOMES, CELSO CAMILO, LEONARDO ARAÚJO LIMA, ALEXANDRE CARDOSO, EDGARD
LAMOUNIER JR., KEIJI YAMANAKA, Artificial Neural Networks to Recognize ARToolKit
Markers, Proc. of Artificial Intelligence and Pattern Recognition, pp. 464-469, 2007.
PAPAGIANNAKIS, G. et al., Mixing Virtual and Real Scenes in the Site of Ancient Pompeii,
Journal Of Computer Animation and Virtual Worlds, December, 2004.
University of Bucharest and Ovidius University of Constanta


316
PARK, Y., and WOO, W., ARTable: AR Based Interaction System Using Tangible Objects, in KCC05,
pp. 523-525, 2005.
POPOVICI, D. M., SEPTSEAULT, C., QUERREC, R., Motivate Them to Communicate, Proceedings
of CW2006, IEEE Computer Society, Geneve, 2006, pp. 198-205, ISBN: 0-7695-2671-3.
VLAHAKIS, V. et al., ARCHEOGUIDE: Challenges and Solutions of a Personalized Augmented Reality
Guide for Archaeological sites, IEEE Computer Graphics and Applications (5) 22: 52-60, 2002.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

317
Online Education Platform: Experior

Andreea Teodorescu, Ciprian Badescu, Radu Ungureanu

Arnia Software, No. 61 Icoanei Street, Bucharest, Romania
E-mail: andreea@arnia.ro


Abstract
In nowadays society, technology has become very complex due to the rapid
worldwide development in all domains of activity. Education has been also
benefiting from this continuous and capacious improvement. Not only has instruction
been converted into a complex system, but, moreover, it gained the capacity of using
technologies, as the Internet, in order to enhance schooling. Therefore, e-learning is
today one of the main areas of interest for the software companies who have in mind
the fact that these new solutions are the future of the educational system. Experior is
an online Mathematics assessment tool addressed to high school students who want
to appraise their Mathematics knowledge level, to train themselves for the
challenging exams they are about to face, to get accustomed to the examination
environment and practice the multiple choice questions tests.

Keywords: assessment, online testing, e-learning platform, Mathematics community,
virtual learning environment.


1. Introduction

Using the Internet in the educational system created the proper conditions for a
completely novel approach in this important field. E-learning is a new perspective on how
school should be organized by the rightful institutions: easy to comprehend with specific
levels, student-centered and efficient.
Arnia Software (www.arnia.ro) developed Experior.ro, an online Mathematics
assessment platform for high school students. By means of its multiple choice questions
tests, its online library containing Mathematics lessons and articles, students can appraise
their Mathematics level, can accumulate knowledge, prepare for exams – be they simple
school tests or bachelor’s exams. Launched in the end of 2007, Experior addresses at this
stage only to the Romanian educational market. Its emphasized interactivity with the user
makes it unique among the educational solutions and tools addressing to the Romanian
instructional segment.


2. Target and characteristics

Experior addresses to Romanian high school students and professors. According to
latest statistics performed on national level, during the educational year 2005/2006, a
University of Bucharest and Ovidius University of Constanta


318
number of 767.439 pupils were enrolled in high school in that particular educational year,
studying in 1410 high school institutions. Some more 284.412 pupils were enrolled in 90
vocational and apprenticeship educational institutions.
Experior offers a variety of features and services, amongst which one can identify:
a complex educational platform comprising of an online testing and assessment module; a
documentation module divided into two components: one dedicated to problems and
solving solutions submission and an online library with access to documentation
materials: Mathematics lessons and articles; an assisted learning module – classes/groups
of students organized within the platform, who can access a study program defined by a
Mathematics professor (homework, lessons to be read and learnt, tests to be taken,
questions and discussions); user evolution statistics; tops and rankings; educational
networking (forum, blog, users' profiles, messaging).


3. A closer look at Experior

Experior is an online educational platform focusing on testing and evaluating
mathematical knowledge. Its main goal is the improvement of students' results in
Mathematics. Either they are interested in the National Educational Program, semester
tests, faculty admission, reviews, national exams or in notable Olympiad performance,
Experior.ro is a complementary tool aiming for the improvement and integration of
students' acquirements.


Platform organization

The team contributing to the e-learning platform consists of content providers,
quality and content assurance – content valuators from a scientific point of view, data
entry operators and proofreaders, as shown below, in figure 1.
The tools defined and used within the platform are: defined template documents for
content providers, word processors for content editing and an internal application
developed for content managing and content packaging, which functions with a pre-
defined word processor.
The platform content is original, created by a team of 6 experienced Mathematics
professors, who are collaborators of Arnia Software. Platform content created by content
providers consists of test units/problems/exercises and documentation materials – articles
and lessons.
The scientific validation of the mathematical content is assured by an experienced
team based on a powerful backend system built by Arnia that permits reviews,
corrections, improvements of the information.
Step by step flow for test units:
• content providers fill in the defined template documents, and constantly deliver
the materials to Arnia;
The 3
rd
International Conference on Virtual Learning, ICVL 2008

319
• materials provided are checked – quality and content assurance step; scientific
validation of content;
• data entry operators (DEO) edit the problems/units in electronic format, by means
of a word processor. Files edited by DEO are packed in problems/test units by means of
the packaging application, and sent to the data server;
• test units/problems are constantly synchronized with the data server;
• an administrator redirects the problems/test units to be approved or to be
corrected, if there are mistakes, to the DEO;
• back to DEO, each operator that has edited problems/test units receives back each
problem package sent, and has to correct/accept/approve the problem;
• a problem/test unit that has been checked and is correct can be published online.

Mathematics Professors Scientific validation & standardization



Experior platform Data entry operators

Figure 1. Informational workflow of the Experior platform

Note: the flow for documentation materials is almost the same as described above for test
units / problems, but significantly simplified in process.

Experior content is defined by over 3000 problems and exercises, more than 200
Mathematics lessons and articles, organized in the online library module, and by over 100 test
categories, structured according to the National Educational Program. The content is weekly
updated by Arnia's data entry operators and a rough approximation reveals numbers like
150 Mathematics problems and exercises and 15 lessons and articles on a weekly basis.


University of Bucharest and Ovidius University of Constanta


320
Platform functionalities

Online testing and assessment module

The online testing and assessment module is based on an education plan split up
into grades, chapters, lessons and knowledge. This division serves to the process of
generating a test from the data base, test that is structured according to parameters like:
time, knowledge coverage, affinity, difficulty, and grade. The multiple choice tests
depend on a certain time to solve, have an exact area of knowledge covered, and provide
access to a detailed solution of each problem/exercise after finishing a test. This option
has the purpose of helping the student understand why his way of dealing with the
problem was wrong or how he could solve the exercise using other methods.
It is worth mentioning that a generated test is not just a set of problems randomly
selected from the database, although a random factor is considered. For example, if a test
is generated to cover a whole year knowledge plan in 180 minutes (recapitulation test),
then the test should cover the large piece of knowledge area with a few problems in such
way that the knowledge coverage and affinity factors are satisfied. For a better
understanding of this fact, let us consider that a knowledge area is made up from
segments as displayed in figure 2, and a problem covers some of these segments.



Figure 2. Segmented display of knowledge coverage factor

The knowledge coverage factor is determined by the red markings on the base
segment representing the knowledge area we need to cover. Each of the problems covers
part of the base segment. The knowledge coverage is the report between the number of
red marked segments and the total amount of segments from the base. Affinity represents
the report between the greater distance between two red marked segments (or one red
marked segment and an end) measured in segments and the total number of segments on
the base. Having clarified this aspect, we conclude that the affinity factor represents the
extent to which the covered segments are scattered on the base. These parameters help to
create a relevant test. If a test should cover a large amount of knowledge and its problems
cover only the beginning of the knowledge area, then the test is irrelevant. After having
finished a test, the application shows the user exactly which knowledge areas were
covered by the test selected problems, as shown in figure 3, in the bottom of the image.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

321



Figure 3. Page with a finalized test, with knowledge area
coverage statistics in the bottom of the page

In the testing and assessment module, the choice of tests is made according to class,
chapter, level of difficulty, and scope. Once the choice made, the user can start taking a test.
The test is generated in a couple of seconds (up to 15 minutes, depending on Internet connection).



Figure 4. New test page versus finalized test page
University of Bucharest and Ovidius University of Constanta


322
Defining elements of a new test page are: predefined time according to goal,
difficulty level, covered knowledge area, timed tests, multiple choice tests, visual guiding
elements. After the final revision of the test, the user declares the test finalized. Defining
elements of a finalized test page are: real-time correction and time assigning, detailed
solving solution; user statistics per knowledge; knowledge coverage rate per test.
Elements of a new test and of a finalized test are shown in figure 4.


Documentation module

The documentation module has a segment which helps students review parts of the
educational curriculum that they didn't fully understand or gives them the chance to learn
new concepts, problems, by reading lessons or by studying articles. All resources are a
handy tool, having in mind the fact that they offer a great variety of examples and
situations students can find in their exercises. Within the library module, the user can
access tests associated to lessons, in order to check and confirm the knowledge covered.
A very important characteristic is that Experior’s online library is well thought-out in
order to ensure a general and also a complete view of the subject matter, and it contains
over 200 Mathematics lessons and articles as shown in figure 5:



Figure 5. Display of the documentation module, online library

The second part of the documentation module is represented by the problems and
solving solutions submission, area within which professors, as well as students, can send
either exercises or solutions to the problems/exercises already submitted. This part is an
interactive, monitored part, which means that each proposed problem, or solving solution
of an existing problem, passes through an inspection, and if it is validated, then it is
written in the agreed internal format, so that the mathematical formulas and text will be
displayed accordingly, as shown in figure 6:

The 3
rd
International Conference on Virtual Learning, ICVL 2008

323


Figure 6. Display of documentation module; problems
and solutions submission


Assisted learning section

Another module developed within Experior is the assisted learning section. This
module recreates, by means of the virtual classrooms, the atmosphere of a typical
classroom. Here, the user has the opportunity to meet two entities: a coordinator (unique)
and the students (as many as the coordinator-professor accepts). The professor can send
some invitations and the students are able to see and accept them if they want, but
students can ask permission to a certain professor to be part of a class, as well. The
classes/groups of students organized within the platform can access a study program
defined by a Mathematics professor, with homework, tests to be taken, lessons and
articles to be studied. Professors are absolute administrators of a class and they can
publish lessons to be read by the students of a class, they can generate tests and verify the
results of each student, and they all have the possibility to use the forum for asking
questions, for giving examples etc. In contrast to other applications, Experior ensures the
entire development of the process online. A quick preview of the professor’s overview of
a virtual classroom is shown in figure 7, displayed below:


University of Bucharest and Ovidius University of Constanta


324



Figure 7. Display of the assisted learning section

A professor can view a student’s test and he can also grade a student based on his
homework or based on the answers provided. All questions and answers are transferred to
the forum; each virtual class gets its private topic, and in that topic a question opens a
thread. Each answer given for a question is transferred inside the question’s thread as a comment.


Statistics module (evolution, graphical statistics)

User evolution statistics display in a graphical, easy-to-understand interface, the
user’s evolution, taking into consideration the time for finishing the test, the difficulty,
and the grade obtained by the user.



The 3
rd
International Conference on Virtual Learning, ICVL 2008

325
The user has access to the graphic representation of the knowledge coverage as
well. Samples for these statistics are shown in the images below:



Figure 8. Statistics – evolution and graphical statistics


Tops and rankings module

Tops and rankings establish a hierarchy based on a score which is calculated using
the grades obtained in tests and the difficulty level. These tops stimulate competition
between students. By means of these tops, students can see how good they are compared
to other students. A top will show a student his rank in his school, and on district/town
level, on a general level, which includes all registered users.


Educational networking module

Another point of interest, especially for students, is the educational networking,
which includes a forum, a blogging platform, users' profiles and possibility to connect
with other registered users within the platform. Main focus here is user interactivity and
providing communication channels in order to facilitate interaction within the
Mathematics community.


Modules in development

Modules of Experior which are now in development: online Mathematics tutoring
module (which has been already implemented, but will be completed with other features);
online Mathematics contests module (expecting an opportunity to be launched).
Arnia is also preparing the extension of the educational plan to the seventh and
eighth degrees, for the testing module and the online library.
University of Bucharest and Ovidius University of Constanta


326
4. Experior promotion campaign

The promotion team, consisting of three promoters trained to make presentations of
the platform, focuses on direct promotion in high-schools, on local (Bucharest) and
national level. So far, the team has covered about 40% of the high-schools in Bucharest,
and the following Romanian cities: Brasov, Iasi, Buzau, Suceava, Targu-Jiu, Braila, and
Vaslui. Feedback of the promotion campaign has been very good so far. Mathematics
professors have shown interest in the solution and also in the possibility to collaborate
with Experior. Good feedback was also received from students, particularly when it
comes to the high interactivity of the solution.


5. Expansion directions

On national level, Arnia targets the implementation of the platform in schools and
high schools (for their Intranet). Other steps would be: the creation of networks between
schools and high schools; the platform extension to other educational curricula areas,
such as: Physics, Economy, foreign languages, appraisal tests; administrative and
governmental projects focusing on education and vocational training; implementation of a
platform for contests and “online Olympiads”.
On international level, Arnia plans the implementation of the platform in other
countries, in partnership or as a financed project; envisaged possibilities: the platform
implementation in universities, private schools and high schools, by means of
collaboration or partnership with these educational institutions, based on the know-how
provided by these institutions; development of an international platform for contests and
“online Olympiads”.


6. Conclusion

Experior key points are: maximum accessibility, being 100% online; access from any
browser, no download or install required; original content; assessment tests; high interactivity
with the user – educational networking; library with Mathematics lessons and articles;
creating a Mathematics community; offering a virtual environment similar to a classroom.



REFERENCES


Internet Sources

http://www.insse.ro/cms/files/pdf/ro/cap8.pdf
http://ec.europa.eu/education/policies/2010/et_2010_en.html
http://www.edu.ro/index.php/articles/9132

The 3
rd
International Conference on Virtual Learning, ICVL 2008

327
Intelligent Systems for Students Knowledge
Automatic Evaluation

Iuliana Dobre
1


(1) Petroleum-Gas University of Ploiesti
39, Bdv. Bucuresti, Ploiesti-100680, ROMANIA
E-mail: iulianadobre@yahoo.com


Abstract
One of the latest world wide trends in regards of educational process improvement is
the developing of new teaching, learning and assessment systems. This trend is
focused on making the new systems more flexible and more efficient using the World
Wide Web capabilities in assisting both, teachers and students. In this paper the
author propose an intelligent system for automatic evaluation of students knowledge
assisted by computers. The ultimate goal pursued by the author is to share the
responsibility of student assessment process between the teachers and students.

Keywords: Knowledge Base, Inference Engine, Intelligent System, Web-based Education.


1. Introduction

During the past few years the Web-based education has influenced educational
process in all aspects of teaching, learning and evaluation by increasing the students’
autonomy and making them to participate actively in instructional decisions and
supporting them to establish more accurate their personal goals and to assess these goals.
At present, the researchers consider that a new education era has begun based on
personalised learning environments. The World Wide Web capabilities and the
personalised learning environments will change deeply the educational process approach
enhancing the widen participations and facilitating the informal and workplace learning
(Magoulas and Chen, 2006).
Both, teachers and students are different. Every teacher teaches and assess
differently. Not every student learns in the same way. Each student brings a different
background, a different experience from the past years of study, and each student has
different expectations. Also, each discipline characteristics and level of material to be
learnt have an influence on educational process.
Many questions arise. How do we teach? How do we learn? How do we assess?
And there are not simple answers to such questions. The actual knowledge about the
relationship between teaching-learning-evaluation is still incomplete, but we do know
enough about the educational process to be able to make fair enough decisions, to design
appropriate learning materials and tests and to establish the actions usually helping in
enabling education to happen.
University of Bucharest and Ovidius University of Constanta


328
Students generally like to have a sense of belonging (Bender, 2003). Therefore, the
author believes that the responsibility of the educational process must be shared with them.


2. Intelligent Systems – Generalities

Generally, an intelligent system has two sub-systems as follows:
Knowledge Base – contains the specific knowledge for the domain, general
knowledge, principles, theories, abstract concepts, ideas etc.;
Inference Engine – is utilizing the knowledge from the Knowledge Base for
problem solving.
Usually, the knowledge is represented through a symbolic form which can be
manipulated by the intelligent system.

The knowledge represents the individual understanding of a specific domain and
can be represented through various types as follows (Alexandru, 2002):
Procedural knowledge – describe how to resolve a specific problem (i.e.: rules,
strategies, procedures);
Declarative knowledge – describe the problem knowledge (i.e.: simple
declarations – false/true, list of declarations which are describing more detailed objects
and concepts);
Meta-knowledge – is the “knowledge” about knowledge (i.e.: strategies,
arguments utilised by experts to increase the system efficiency);
Heuristic knowledge – arguments (i.e.: empirical arguments/strategies obtained
from experiments);
Structural knowledge – describe various knowledge structures (i.e.: concepts,
sub-concepts, objects).
The Knowledge Base includes implicit and explicit knowledge. Due to this a
Knowledge Base is different than a common data base.


3. An Intelligent System for Students Knowledge Evaluation

An assessment process, generally speaking, must be a fair process giving equal
chances to all students. Ensuring the maximum grade of process fairness could be the
very first step in the direction of making students more responsible by getting their trust
in the teacher “balance of appreciation” (Dobre, 2007).
This intelligent system has designed and developed like an integrated system for
learning and evaluation addressed to Informatics discipline, having as general
functionality principle the assisting of learning and evaluation processes by computers.
The system can be used within internal networks (laboratories networks) as well as
through Web browsers.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

329
The system was implemented/tested with “Procedural Programming” course part of
Informatics discipline programme. The course has been structured in ten theoretical
chapters and ten application laboratories. Also, has been considered for the evaluation
process the formative and summative evaluations of students’ knowledge. The author has
chosen like knowledge representation method the semantic networks.
This technique has been selected in order to control with effectiveness the students’
status in any moment and at any level during the learning process assisted by computer
and in order to generate electronically the courses based on some courses already existed
in a data base.
A semantic network is a graphical representation of a graph as follows:

[1] ( ) U X G , =

where:

X = is the set of nodes representing the primitive elements for concepts, events, states;
U = is the set of arcs representing the primitive elements which represents the
abstraction of relations between concepts.

The semantic networks have three main elements as follows:
Concept – any idea with a meaning and having a unique label or definition;
Relation – represents the connection between two concepts;
Instance – represents an event of two concepts allied between relations.

The main relations specific to a semantic network are (Oprea, 2005):
1. Individual-Generic Network or Instance-Class – is the relation ISA (the
abbreviation is coming from “Is a”, English language), to represent the membership of
one object to a set;
2. Generic-Generic Relation or Sub-Class-Sub-Class – is the relation AKO (the
abbreviation is coming from “A kind of”, English language);
3. SUBSET A;
4. HAS-PARTS;
5. AGENT;
6. OBJECT;
7. ATRIBUTE.


3.1. Example of a Semantic Network

Below the author presents as example of a semantic network associated to the
instructional objective defined as “Program Control Instructions”, figure no. 1. The
University of Bucharest and Ovidius University of Constanta


330
example is based on the use of AKO relation and other particular types of relations (i.e.:
general form, organigram, effect etc.).

3.2. The General Diagram of the Proposed System

The Inference Engine is using inference general or particular rules in order to arrive
to those conclusions which are not represented in the Knowledge Base. In figure no. 2 is
presented the general diagram of the system proposed by the author. From all known
inference strategies the author has chosen the chaining forward strategy. The utilised
strategy starts from a known set of facts and obtained new facts utilising the conclusions
of the rules with premises matching the initial known facts.
The process is continued until the final scope is achieved or until there are no more
rules with premises matching the initial known facts or with derivate facts during the
inference process.
The working memory will be formed from the facts provided by user and the facts
provided by system. The Knowledge Base of the proposed system have, in the same time,
files specifics to data bases containing the questions from the assessment tests used to
achieve the instructional objectives.

Program Control
Instructions
Expression
Instructions
Decisional
Instructions
Iteration
Instructions
Jump
Instructions
Instruction
“while”
Instruction
“for”
Instruction
“do-while”
A
K
O
A
K
O
A
K
O
A
K
O
A
K
O
A
K
O
A
K
O
while (condition)
instruction;
G
e
n
e
r
a
l F
o
r
m
E
f
f
e
c
t
The instruction is
executed as long as it
is true (not equal with
0). The program
execution is continued
with the next
instruction.
Condition
Instruction
True
False
O
r
g
a
n
i
g
r
a
m
If from START the
condition is false the
instruction will not be
executed
O
b
s
e
rv
a
tio
n
s








Figure 1. Semantic network associated to the instructional
objective defined as “Program Control Instructions”
The 3
rd
International Conference on Virtual Learning, ICVL 2008

331

Module 1
Evaluation
Learning
Module 1
System Inputs
LEARNING SUBSYSTEM
Aplications
Module 1
Number of
runs through
Module 1
(evaluation=0)
Satisfactory
results
Unsatisfactory
results
Number of
runs through
Module 1 de
(evaluation ++)
Number
of runs through
Module
1<=2
YES
NO
EVALUATION SUBSYSTEM
Module N
Evaluation
Aplications
Module N
Number of
runs through
Module N
(evaluation=0)
Satisfactory
results
Unsatisfactory
results
Number of
runs through
Module N
(evaluation++)
Number
of runs through
Module
N<=2
YES NO
Learning
Module N
Evaluation
Instructors
Evaluation
System
Evaluation
Students
Knowledge
Evaluation
System Outputs
To the next learning cycle
inputs
Next learning
cycle inputs
The achieved knowledge
and skills quantum as per
curriculum
Students Groups
Instructors
Timing schedule of
the learning & evaluation
processes
Logistic preparation
referring to the learning
assisted by computers
(availability of equipments ,
courses etc.)
The list with students results
divided by phases and
modules
Results histogram
Conclusions about successes
and failures
Global conclusions about
successes and failures


Figure 2. The general diagram of an intelligent system
for students’ knowledge automatic evaluation


3.3. The Algorithm Used to Calculate the Final Note for Each Student

The Algorithm proposed by author to calculate the final note for each student is
described with the below pseudocode [2].


University of Bucharest and Ovidius University of Constanta


332


.
; / ] [
]; [ ] [
1 , 1 , 1
; 0
; 4 1
; 1
; 0 ] [
; 1
; 1 ] [
5 ] [ 3
; 0 ] [
; 10 / ] [
]; [
]; [ ] [
1 ], [ , 1
; 0
; 0
;
;
;
5 ] [ 3
; 0 ] [
; 1
; 1
1 , , 1
Endfor
N sum s Notaf
Endfor
j Ok j Nota sum sum
do i j For
sum
k or N i Until
Endwhile
Endif
k k
i Nota
Else
i i
i Ok
then i Nota and k If
i Ok
B A i Nota
Endfor
j gr B B
j pct j gr A A
do i nr j For
B
A
i evaluare Nodul Run
i aplicatii Modul Run
i predare Modul Run
do i Nota and k While
i Nota
k
repeat
i
do nrs s For
=
∗ + =
+ − =
=
= + =
+ =
=
+ =
=
≥ ≤
=
∗ =
+ =
∗ + =
+ =
=
=
< ≤
=
=
=
+ =


In the above algorithm [2] the author has used the following annotations:

[2]
The 3
rd
International Conference on Virtual Learning, ICVL 2008

333
N = objectives number;
Nota[i] = the result obtained after module i assessment;
nrs = the number of students assessed;
gr[i] = the questions difficulty grade associated to a question;
pct[j] = takes value 1 if the student answer to question j is correct and 0 if the
answer is not correct;
Notaf[s] = the final note of a student, s.

Questions difficulty grade was established on a range from 1 to 5. The Knowledge
Base for each course instructional objective were uploaded a set of questions, covering
each learning module. The questions are chosen randomly in a defined number. The
number of questions is established by the knowledge engineer or expert.
Using the chaining forward strategy the working memory has been enriched by
calculating for each question the following ratio:

[3]
] [ .
] [ .
] [
i question to answered who students of No
i question to correctly answered who students of No
i r =

The rules proposed by the author to enrich the system Knowledge Base were the
following:

[4]
. 1 ] [ ] [
4 ] [
9 , 0 ] [ :
1
+ =


i gr i gr THEN
i gr AND
i r IF R


[5]
. 1 ] [ ] [
1 ] [
1 , 0 ] [ :
2
− =
>

i gr i gr THEN
i gr AND
i r IF R


This system has been built using the following software resources: PHP server-side
scripting language (is combining the Perl, Java and C concepts), MySQL (Structured
Query Language) and the Web Server Apache HTTP.
The feedback regulator of the proposed system determines the command in a way
that after the process is resumed the system is achieving the target which is the objective
evaluation of students’ knowledge to a specific discipline.


4. Conclusions

The present time requirements in higher educational process involve the re-ordering
in the teachers’ ability to create, acquire, assimilate and share the knowledge to and with
University of Bucharest and Ovidius University of Constanta


334
their students. Knowledge sharing methods and techniques will be improved significantly
in the next short term future and all these transformation, in the author opinion, will leave
no room to another option than to follow this mega-trend.
Development and continual improvement of the knowledge base at individual level,
for each student, independently but in same time without moving outside from the
modern society requirements is the next major challenge for the higher education.
Today the knowledge is one of the key resources for the human society and the
higher education institutions are playing a critical role in developing this valuable asset
(Bender, 2003). The author believes that integrating learning and assessment by sharing
the responsibility of the educational process goals achievement is the proper solution for
these incoming transformations.
Also, the author believes that a true value assessment of the student learning begins
always with the educational values. As more and more institutions from higher education
environment incorporate online courses in their curriculum, the teachers need to
determine and implement better methodologies and techniques for students’ knowledge
evaluation. The system proposed is just another solution trying to integrate successfully a
web-based education integrated system for learning and assessment with the today
requirements in higher education environments.



REFERENCES



Books

ALEXANDRU, A. (2002), Sisteme Expert – Concepte şi AplicaŃii, Matrix Rom, Bucureşti.
BENDER, T. (2003), Discussion – Based Online Teaching to Enhance Student Learning, Stylus Publishing
LLC, Virginia.
MAGOULAS, G. and CHEN, S. (2006), Advances in Web-Based Education: Personalized Learning
Environments, Information Science Publishing, London.
OPREA, M. and NICOARĂ, S. (2005), InteligenŃă Artificială, Editura UPG, Ploieşti.


Journal Articles

Dobre, I. (2007) Evaluation of Students Knowledge – An Experiment in E-Learning. Buletinul UniversităŃii
Petrol-Gaze din Ploieşti LIX, 2, 43-48.


The 3
rd
International Conference on Virtual Learning, ICVL 2008

335
Timetable Planning using Intelligent Agents

Irina Tudor
1
, Mădălina Cărbureanu
1


(1) Department of Informatics, Petroleum-Gas University of Ploieşti,
39, Bd. Bucureşti, 100680, ROMÂNIA
E-mail: tirinelle@yahoo.com


Abstract
At the beginning of each semester a frequent problem appears regarding the
timetable planning. It is not an easy task for the person responsible with solving this
problem because he/she must take into consideration many constrains such as:
teaching stuff timetable options, elaboration rules (number of lectures, number of
seminars, number of practical activities per day, etc.) and the number of students
which determines the allocated lecture/seminar hall. In this case, an adequate
method for reducing the execution effort and time is the designing a multi-agent
system (MAS) using ZEUS software, developed by British Telecom. Our work
consists in identifying and implementing the necessary agents with their tasks for
sharing the resources, establishing the communication ontology and coordination.
This paper highlights the opportunity of multi-agent systems application in the
superior education field.

Keywords: Intelligent Agent, Ontology, Multi-Agent System, Education.


1. Introduction

In recent years, a major research effort in artificial intelligence domain has been invested in
designing and building intelligent agents-software or hardware entities that interact with an
external environment in an intelligent way.
In computer science, an intelligent agent (IA) is a software agent that assists users and acts
on their behalf, in performing non-repetitive computer-related tasks, in the sense of a
representative agent. Intelligent agents are used for operator assistance or data mining. The
intelligence implies the ability to adapt and learn (Wikipedia, Software agents, 2008).
In artificial intelligence, an agent is used for intelligent actors that observe and act upon an
environment, in the sense of a rational agent.
In our paper we present an application of intelligent agents in the superior education field.
A solution for timetable planning in the framework of a department is given.


2. About Intelligent Agents

In literature, an agent is known as an entity that perceives, reasons and acts. In
computational terms, that which is perceived is an input, to reason is to compute, to act is to output
the result of computation. An agent is equipped with objectives and the rational quality consists in
acting optimally with respect to its objectives.
University of Bucharest and Ovidius University of Constanta


336
Intelligent agents perform a wide range of services, including automatic searching,
answering specific questions, providing information and updates about events, running programs
and presentations, reporting current news, comparison shopping, and tutoring (Baylor, 1999). This
technology combines artificial intelligence (reasoning, planning, natural language processing, etc.)
and system development techniques (object-oriented programming, scripting languages, human-
machine interface, distributed processing, etc.).
An intelligent agent can be used to perform various activities such as: searching for
information automatically, answering to the specific questions, informing users when an event has
occurred, providing custom news to user on a just-in-time format, intelligent tutoring, automatic
services, such as checking web pages for changes or broken links.
An intelligent agent can be applied successfully in various fields, as follow: workflow
management, network management, air-traffic control, business process engineering, command
and control, education, digital libraries, information management, data mining, electronic commerce.
An agent runs if two conditions are met. The former is a common language, called Agent
Communication Language (ACL), that must exist in order to enable software to recognize the
intention behind a request of an agent and, as a latter condition, there must be an architecture,
where a piece of software can describe its abilities and needs.
Various special languages have been developed to facilitate the communication between
agents and the most common are: FIFA ACL, Actor, Tcl/Tk, Telescript, Linda (mostly for mobile
agents), Agent0, Concurrent Metaterm, KQML, etc. Communication protocol is not a low-level
protocol but a protocol establishing possible actions of agents in every moment of communication
with other entities.
A FIPA ACL message contains a set of one or more message parameters. Precisely which
parameters are needed for effective agent communication will vary according to the situation. The
only parameter that is mandatory in all ACL messages is the performative one, although it is
expected that most ACL messages will also contain sender, receiver and content parameters. If an
agent does not recognize or is unable to process one or more of the parameters or parameter
values, it can reply with the appropriate not-understood message (FIPA Abstract Architecture
Specification. Foundation for Intelligent Physical Agents, 2000).


3. Multi-Agent System Design

In this application there are defined four agents, capable to interact in order to minimize the
effort and time for timetable planning process.
The above-mentioned agents are the following: Timetable administrator, Timetable teacher,
Environment and Restriction. Also, one may use utility agents that are automatically generated by
Zeus: Broker, Visualiser and ANS (Agent Name Server).
Various roles and tasks are associated to each agent. For instance, the functions for
Timetable_admin agent are:
1. Requesting the teachers their timetable options;
2. A new timetable option solicitation, when the current option is not valid;
3. The timetable supplying;
4. The supplying of the laboratories loading.
For the Timetable_teacher agent, the associated functions are:
1. Verifying the laboratory’s state;
2. Delivering laboratory activities;
3. Sending the solicited timetable option;
4. Sending a new timetable option.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

337
The role of the Environment agent is the laboratory information supplying and the only role
of the Restriction agent is to assure restrictions application.
For having a good image upon the information flux, the representation of the information
flux and interaction diagram is necessary (figure 1, figure 2).



Figure 1. Information flux



Figure 2. Interaction diagram

University of Bucharest and Ovidius University of Constanta


338
An agent PAGE (Perception, Actions, Goals and Environment) description consists in the
perceptions, actions, goals and environment (in which the agent interacts with other agent/agents)
identification. The table below displays the PAGE description for the proposed agents in our application:

Table 1
PAGE description

Agent Perception Action Goal Environment
Timetable_admin

Receiving
options
Asking
options
Generating
timetable
Laboratory
Timetable_Teacher

Receiving tasks
Answering
tasks
Timetable
establishing
Laboratory
Environment

Laboratory
state
Supplying
laboratory
information
Supplying
laboratory
information
Laboratory
Restriction


Restrictions
applying
Restrictions
discharging
Laboratory


4. The Multi-Agent System Implementation with Zeus Agent Toolkit

An essential step in MAS developing process is the knowledge modelling, consisting in the
ontology representation. In the context of computer and information sciences, ontology defines a
set of representational primitives with which one can model a domain of knowledge or discourse.
The representational primitives are typically classes (or sets), attributes (or properties), and
relationships (or relations among class members).
The definitions of the representational primitives include information about their meaning
and constraints on their logically consistent application (Gruber, 2008).
In our application the agent’s ontology is composed from the terms used in the
communication process. The figure below presents the developed ontology:



Figure 3. Agent ontology

The 3
rd
International Conference on Virtual Learning, ICVL 2008

339
The next step consists in the agent’s development and identification of associated tasks. The
development of agents for our application is presented in the figure below:



Figure 4. The agents in Zeus

The development of an agent, for instance Timetable_admin, implies the realization of the
next intermediary steps: the Zeus implementation of the agent tasks, establishing of the relation of
the agent with other agents (peer to peer, subordinate or superiority relation), establishing the
coordination protocols (for the respondent or initiator agent role), allocating the initial resources
and establishing the maximum number of simultaneous tasks. All these are established in the
following panels: Agent Definition Panel, Agent Organization Panel, Agent Coordination Panel
and Value Restriction Panel.
The implementation of the Supply_timetable task associated with Timetable_admin agent is
presented in the next figure:

University of Bucharest and Ovidius University of Constanta


340


Figure 5. Task Preconditions and Effects

Zeus software provides a batch of utility agents: Agent Name Server (ANS), Facilitator and
Visualiser. These agents are implicitly generated by Zeus and they form, beside the user-developed
agents, the application agent society. The resulted agents society is presented in the figure below:



Figure 6. Agents society

Using the proposed ontology and the ACL as a communication language (Agent
Communication Language) for reaching the application goal, the agents start to change messages
(i.e. sending call for proposals (cfp), sending proposals, proposals’ acceptance/refusal, sending
information, etc.),as presented in the figure:

The 3
rd
International Conference on Virtual Learning, ICVL 2008

341


Figure 7. Interactions between agents

As a result of agents communication a timetable is generated in a primary form represented
by attributes with values which can be interpreted to obtain a useful timetable form.



Figure 8. The results table

University of Bucharest and Ovidius University of Constanta


342
REFERENCES


FIPA Abstract Architecture Specification, Foundation for Intelligent Physical Agents, 2000,
http://www.fipa.org/specs/fipa00001
BAYLOR, A. L. (1999), Intelligent Agents as Cognitive Tools, Educational Technology, 39(2),
36-40.
COLLINS, J., NDUMU, D. (1999), The Application Realisation Guide, Intelligent Systems
Research Group, BT Labs.
COLLINS, J., NDUMU, D. (1999), ZEUS Technical Manual, Intelligent Systems Research Group,
BT Labs, 1999.
COLLINS, J., NDUMU, D. (1999), The Role Modelling Guide, Intelligent Systems Research
Group, BT Labs.
GRUBER, T. (2008), Ontology to Appear in Encyclopedia of Database Systems, Ling Liu and M.
Tamer Özsu (eds.), Springer-Verlag, 2008, http://tomgruber.org/writing/ontology-
definition-2007.htm
Wikipedia, Software agents, http://en.wikipedia.org/wiki/Software_agents


The 3
rd
International Conference on Virtual Learning, ICVL 2008

343
Hypermedia System for Online e-Learning and
e-Testing in Project Management

Eugen Zaharescu
1
, Georgeta-Atena Zaharescu
2


(1) "OVIDIUS" University of Constanta
124, Mamaia Blv., Constanta 900527, ROMANIA
E-mail: ezaharescu@univ-ovidius.ro
(2) "DECEBAL" High School, Constanta


Abstract
The main purpose of this paper is to present a flexible hypermedia system for online
e-learning and e-testing in the domain of project management education. This virtual
learning environment uses the newest open source web technologies available for the
moment: PHP, MySQL, AJAX, XML and XSLT. The paper describes the implementation
principles of this e-education system structured on two functional sections:
hypermedia-based e-learning and online dynamically adapted e-testing. In the first
section, a powerful database system allows online access to a very well organized
video tutorial, covering the main aspects of project management e-learning. In the
second section, tests are automatically generated and adapted for each student as
they are presented as shuffled sequence of questions and alternative answers, using a
web-based interface.

Keywords: e-Learning, e-Testing, Distance Learning, Hypermedia, Computer-Assisted
Education, Educational Platform, Virtual Learning Environment.


1. Introduction

The hypermedia system for project management education, presented in this paper belongs
to the wide domain of e-learning and e-testing (or e-assessment). The most important feature of
this modern education domain is the emergence of the ICT (Information and Communication
Technologies) use in knowledge production, diffusion, consultation and automatic assessment. The
generalization of the use of ICT in e-learning, leads to an explosion of LU (Learning Units) on the
internet. Indeed, many studies [3; 6; 7] reveal hundreds of LMS (Learning Management System)
able to provide LU for e-learning, but these are not always reusable by other LMS.
In the last decade, two approaches have tried to answer to this problem of LU reuse. The
first approach is to create repositories of LU shared on internet like research projects ARIADNE,
COLIS, Edusource, DLESE and MERLOT. The second approach is to reuse the educational
scenario as a whole.
Also, new educational languages, standards and specifications, like IMS (Instructional
Management Systems), EML (Educational Modelling Language) and MISA (Méthode d’Ingénierie
des Systèmes d’Apprentissage), propose models for educational scenarios design and reuse.
The emergence of the ICT leads to an explosion of the web based tools and services
(forum, chat, LMS, etc.) which are not always interoperable. The new web technologies and W3C
recommendations represent a solution for the interoperability of these tools and services and the
development of a virtual exporting/importing space.
University of Bucharest and Ovidius University of Constanta


344
The architecture of our hypermedia system for project management education is structured
in two sections:
1. The first section is a LMS (Learning Management System), based on a set of web
services. The hypermedia catalogue developed in this section is able to share the HLU
(Hypermedia Learning Units) located in local or distant hypermedia repositories.
2. The second section is an AMS (Assessment Management System) based on a dynamic
tests generator and a statistical evaluator.
This proposed e-learning and e-testing (e-assessment) management system can perform
automatic test generation and grading from a wide variety of question types supported. It is a data
driven system which can dynamically adjust e-test contents and parameters and fast reorder the
whole e-test form by random number generation.
The structure of this paper is as follows: section 2 presents the concepts and principles upon
learning management system is based on; section 3 describes the functional architecture of
learning management system; section 4 shortly presents the implemented learning unit catalogue
and repository; section 5 presents the concepts and principles of this web based e-testing system;
section 6 describes Web based e-testing system architecture; in the end the experimental and
statistical results followed by final conclusions are revealed in section 7 and 8.


2. Learning Management System. Concepts and Principles

The modelling language, proposed by the IMS (Instructional Management Systems) Global
Learning Consortium is widely inspired from R. Koper works [2]. It provides a rich terminology
which allows to describe in a formal way and to implement reusable educational scenarios. Also, it
offers an educational flexibility because the designer can describe every type of LU (Learning
Unit) (e.g. lessons, problem based learning, etc.).
An LU is introduced by IMS (Instructional Management Systems) Global Learning
Consortium as an abstract item which makes reference to an element of learning or education as
for example a lesson or a module [1]. It is to note that an LU represents more than an orderly
collection of resources; it also includes a variety of prescribed activities (e.g. search activities,
evaluation activities, training activities, etc.), the services, the tools and the resources produced by
the learners and the staff.
The activities, the roles, the resources and the workflow depend of the ones from the others
in the educational scenario.
Conceptually, an LU is modelled as a content package containing the educational scenario.
The content of LU is built according to the IMS content package. It is composed of the following
two major components [1]:
1. The first important component of LU is the manifest which describes the content
structure and the associated resources. It is an XML (eXtensible Markup Language) file, called
“imsmanifest”. The element <manifest> is the root of the manifest file. It contains three direct
children elements:
1.1. The first child is an optional element, called metadata. It describes the manifest as a
whole and uses the IEEE-LOM [4] metadata scheme.
1.2. The second child is called organizations. It describes how the content is organized to
be delivered to the learners. To create the educational scenario for the LU, the <organizations>
element includes the <learningdesign> element. This last one contains the elements which
describe the educational scenario. It summarizes the idea according to which the educational
scenario takes place as a theatre play. The educational scenario is organized in acts in which the
activities are proposed to the roles in a computer environment consisted of learning objects and of
services (chat, forum, e-mail, etc.). It is designed to allow reaching the learning objectives. It is
described according to the hypothesis of some prerequisites which a learner must have to realize
the activity. The educational scenario is organized in A, B and C three levels [1]. The level A is
constituted by the general description elements of the educational scenario. While the B level, adds
to the A level, the elements of the educational scenario personalization (conditions and properties).
Finally, the level C, adds the notification mechanism which allows making dynamic the
educational scenario.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

345
1.3. The last child is called resources. It is a collection of references to resources. The
element <resources> consists of several (zero or more) <resource> element. A resource is not
necessarily composed by a single file. It can be also constituted by a set of files. Each file of
<resource> element is represented by <file> element. These files can be internal files referenced
by relative address or external files referenced by URL (Uniform Resource Locators).
2. The second important component of LU is represented by the physical files (local or
external files) associated to the very contents of LU, itself. They are electronic representations of
media, such as text, sound, images, animations, graphs or any piece of data that can be rendered
from the Internet and presented to learning subjects.
Each of these media may have multiple digital formats (e.g. WAV, MP3 or WMA for
sound files, .AVI, .WMV, .MPG or .SWF for video files). A physical file can be created by the
LU designer or reused from a repository.
The internal files must be included inside the PIF (Package Interchange File) file. The
manifest file and all other XML control files (DTD, XSD) identified by the manifest must be
placed at the root of the PIF file, which is a concise web delivery zip format. The use of a concise
zip format facilitates and accelerates the transport of the PIF file over the Internet.


3. Functional Architecture of Learning Management System

The most important architectural elements of Learning Management System are:
• the actors representing the persons who play the various administrative, educational,
technical roles.
• the LMS representing the learning management and the organization sub-system or core
system. It will realise management of the learners, individualization of the learning, evaluation of
the learners, etc.
• the CMS (Content Management System) representing the LU management system. It
helps to create, to updates and to manage the LU.
These two sub-systems are based on two principles. The first principle is the separation of
the contents and the form. It allows the designers to concentrate on the design and the creation of
contents without worrying by the form. Some CMS proposes predefined models which the
designer uses to insert their contents. The contents consist from the existing resources (reuse) or
created from the new resources. The second principle is the import and the export of the LU. A
LCMS offers both LMS and CMS combined functionalities (LCMS = LMS + CMS).
• The LU repositories are data bases containing LU. They also implement web services
which allow their interoperability with the catalogue, the LMS and the CMS.
• The catalogue is the tool which allows sharing of the LU on the network. Also, it allows
searching the LU on LU repositories according to some search criteria. The LU which answers to
search criteria is downloaded from the LU repositories. They are then used by the CMS or by the LMS.



Figure 1. The functional architecture of LMS (Learning Management System)
University of Bucharest and Ovidius University of Constanta


346
4. Learning Unit Catalogue and Repository

There are two types of LU repositories: local and distant. The local repository is a data base which
is on the system server and contains the LU created by actors. They can be imported or modified
by LMS or CMS. The distant LU repository is a data base containing LU located on another web
server. The LU belonging to this data base is downloadable by the LU catalogue module, only.



Figure 2. Learning unit catalogue and repository during training process


5. Web Based e-Testing System. Concepts and Principles

Definition 1

Question bank is the core of every e-testing system and is represented by a database of
unique questions with parameters, from which the test generation module will make simple selections.
The necessary question parameters may be:
• question type according to possible answer/answers;
• question weight/value for final summarized grading;
• question domain/area/section following the theory classification etc.
The architecture of e-testing systems should have the following modules:
• question input module with special forms for question entry process;
• question importing module from other similar e-testing systems database;
• question removing module from own database.
In order to create a an exchanging space for importing/exporting questions and to provide
compatibility between different systems, several standard structuring forms have been developed
for the elements from question banks. Among them, the most important and promising standard is
developed by IMS (Instructional Management Systems) Global Learning Consortium, Inc.
http://www.imsglobal.org [1]. These system standard specifications are defined in XML, following
W3C Consortium recommendations. The IMS Question & Test Interoperability (QTI)
specification describes a data model for the representation of questions and test data and their
corresponding results reports. Since 2005 starting with version 2.0, QTI supports parameterized
questions via assessment item templates [1].
The e-testing system question bank may have two different types of questions:
• fixed answer question (objective question) and
• free answer question (unobjective question).
The 3
rd
International Conference on Virtual Learning, ICVL 2008

347
The fixed answer question (objective question) is made up of a text body for problem
description and a list of possible answers, where the student must choose from, the correct
one/ones. Most of the e-testing systems use this type of questions in assessment process.
The taxonomy of fixed answer questions may be subsequently defined like this:
• multiple answer question is the most used type in automated assessment.
• short answer question requires an short text answer to be provided;
• short text or numerical value answer question requires a computed result;
• hot-spot question or visual/interactive answer question requires an object/position
identification, graphical element connections, etc.
Furthermore, for multiple answer question category we may have several variations:
• Yes/No or True/False answer question has only two opposite alternative answers;
• MC/SA(Multiple-Choice/Single-Answer) question has only one correct answer;
• MC/MA(Multiple-Choice/Multiple-Answer) question has more correct answers;
• Priority Setting/Selection answer question require items ranking.
The free answer questions (unobjective questions) have no predefined answer. They are
usually used when assessing higher levels of learning domains or Bloom's taxonomy:
Cognitive(Knowledge)-mental skills, Affective(Attitude)-growth in feelings or emotional areas
and Psychomotor(Skills)-manual or physical skills. The free answer question (unobjective
question) category may be divided in two sub-types/sub-categories: program code answer
question and essay answer question.
In the end, it may be concluded that using question banks instead of static test creation
method, we obtain significant advantages in e-testing system:
• a wide set of possible learning objects to choose from in assessing process;
• test generation time is decreased very much.


Definition 2

Test creation algorithm represents the questions selection process from the question bank
(system core), followed by the generation of student’s presentation form. There are 5 different test
delivery models [8], depending on the characteristics of the tested subject knowledge: linear,
dynamic linear, testlets, mastery models, adaptative:



Figure 3. Test delivery models

Linear tests are not adaptive to users and consist of predefined questions and predefined
order. Assessment can be done automatically and results be summarized.
Adaptive tests depend on student’s knowledge. The parameters of test generation are
defined dynamically during the test according to given answers.


Definition 3

Grading and results reporting are the final actions of the automatic e-testing system
which will display the results of the assessment immediately after the end of assessment process,
when all answers of all questions are definitely entered.
The evaluation of the entered answers can be made in two different moments of time: at the
end of the entire test if the system let the subject the possibility to change the entered answers or
after each answer is entered otherwise.
University of Bucharest and Ovidius University of Constanta


348
6. Web Based e-Testing System Architecture

The design approach of TestManager system is based upon a three layer architecture:
database layer (it stores the question bank on a database server and communicates with web
server to generate dynamic web pages), application layer (it receives requests from user interface
layer and generates appropriate answer to every user request; it is executed on web server) and
user interface layer (used by users to submit requests through application logic layer; it is
executed on every client station).



Figure 4. E-testing system question bank and database tables structure (database layer)





Figure 5. The three layer architecture of TestManager system



Figure 6. Assessment process and final statistical results reports (user interface layer)
User interface layer

Application layer

Database layer

The 3
rd
International Conference on Virtual Learning, ICVL 2008

349
7. Experimental and Statistical Results

The e-testing system implements automatic evaluation at the end of the test, when the
knowledge tested subject finalize himself the answer entry process. The system displays subject
final results compared with the complete set of correct answers.
We have implemented negative grading, in order to eliminate intelligent or lucky guessing.
So, the number of points given for any true answer selection (m
T
) and the number of points taken
for any false answer selection (m
F
) will be given by the expressions:

[1]
T
T
n
m
m = ,
) (
T A
F
n n
m
m

− =

where n
A
= total number of possible answers, n
T
= total number of true answers and m =
maximum grade or total number of test points. The final calculated grade will be:

[2]
) (
* ) ( * * ) ( * ) (
T A
T T
T
T F T T T T T
n n
m
n N
n
m
N m N n m N N M

− + = − + =

where N
T
= total number of true answer selections.
The final grade of every test will be displayed as either, absolute value and relative value
(percentage of possible points).
We have repeatedly evaluated a sample of 10 subjects with randomly generated tests using
both normal and negative grading. Number of incorrectly passed test by intelligent or lucky
guessing, instead of real knowledge, tends to lower with the increasing number of questions per
test and with the negative grading, as shown in the table 1:

Table 1
Number of incorrectly passed test by intelligent or lucky guessing

Number of questions / test
Grading Type
40 questions 50 questions 60 questions
Normal Grading 9 % 6.7 % 4.5 %
Negative Grading 4 % 1.7 % 0.9 %


8. Conclusions

This paper presents and analyses the main features of TestManager, a hypermedia system
for e-learning and e-testing dedicated to project managers. The design of TestManager conforms
to the models and specifications provided by IMS (Instructional Management Systems) Global
Learning Consortium.
Therefore, the system can be easily adapted to any education domain because it produces
reusable LU (Learning Units). The Web based e-assessment section of TestManager has a data
driven system design as the main concept behind this system is the question bank. In fact, this
represents the system core, where questions are selected from, during test generation process. The
test creation algorithm dynamically generates equally weighted tests according to previously
predefined strategy. Finally, the negative grading method statistically eliminates intelligent or
lucky guessing answers.


University of Bucharest and Ovidius University of Constanta


350
REFERENCES


[1] IMS Content Packaging Information Model, Learning Design Information Model:
http://www.imsglobal.org/content/packaging http://www.imsglobal.org/learningdesign
[2] KOPER, R. (2005), Modelling Units of Study from a Pedagogical Perspective, the pedagogical
metamodel behind EML, http://eml.ou.nl/introduction/docs/ped-metadodel.pdf.
[3] LMS Open Source, https://fnl.ch/LOBs/LOs_Public/OpenSourcePlatf.htm
[4] LOM metadata official web page, http://ltsc.ieee.org/wg12/files/LOM_Final.pdf
[5] MERRILL, M.-D. (1994), Principles of Instructional Design, New Jersey Educational
Technology Publications.
[6] Oravep Study, http://www.educnet.education.fr/superieur/plateforme.htm
[7] Thot official web page: http://thot.cursus.edu/
[8] PATELIS, T. (2006), An overview of computer-based testing. The college board:
http://www.collegeboard.com/research/html/rn09.pdf
[9] PETTIGREW, M. (2001), Random guessing on multiple choice tests, http://www.shu.ac.
uk/services/lti/people/mp/mcq/
[10] Web Page of LORNET project, http://www.lornet.org/index.htm

The 3
rd
International Conference on Virtual Learning, ICVL 2008

351
The Miracle of the Age: Internet in classrooms

Gülen Onurkan Aliusta, Zehra Unveren, Fatma Basri

Eastern Mediterranean University, English Preparatory School, North Cyprus
E-mail: gulen.onurkan@emu.edu.tr


Abstract
Today, web-based projects are widely used in most language learning institutions. In
web-based projects, students are usually required to do research on a given topic by
using the Internet in order to accomplish a given task. In EMUEPS, students are
assigned web-based projects for their portfolios which constitute a part of the
continuous assessment system. Dealing with web-based projects seems to be hard
work for students as they struggle with using a computer and also the Internet
resources at the same time. Teachers also have hard times trying to help students use
the Internet resources effectively and efficiently and also to evaluate and give
feedback to their students’ project drafts and final copies. The main aim of this
research study was to investigate students’ perceptions of web-based projects. The
findings of this study looked for improvement in English language learning as a
result of integrating web-based projects in the curriculum. 60 students participated
in this study. Two sets of data were collected: Internet use survey and student
interviews. The results gathered both from the interviews and the survey revealed a
positive attitude towards the use of the Internet for portfolio projects.

Keywords: Internet, Web-based projects, Language learning.


1. Introduction

Today, web-based projects are widely used in most language learning institutions. In web-
based projects, students are usually required to do research on a given topic by using the Internet
in order to accomplish a given task. “The procedure for educating students has shifted from
providing them with information to opening doors for them to explore topics and to create
meaningful learning experiences for themselves” (Smaldino, S, E. and et al, 2005, p. 118).
In English Preparatory School of the Eastern Mediterranean University (EMUEPS) in North
Cyprus, students are assigned web-based projects for their portfolios which constitute a part of the
continuous assessment system. The EPS has integrated web-based projects into the curriculum in
order to enable the students to use the web resources effectively and efficiently for their academic
studies and also help them improve their English language.


2. Statement of the Problem

Dealing with web-based projects seems to be hard work for students as they struggle with
using a computer and also the Internet resources at the same time. For example, students who have
limited previous experience with the web either find it difficult to retrieve the information they
University of Bucharest and Ovidius University of Constanta


352
need or can make no sense of the results of the search. The teachers also have hard times trying to
help students use the Internet resources effectively and efficiently and also to evaluate and give
feedback to their students’ project drafts and final copies. Therefore, using the web requires
additional effort from both parties.


3. Aim of the Research

The teachers and the administrators at the EPS need to know students’ perceptions of using
the Internet for their portfolio projects because students’ attitudes towards the Internet directly
affect their motivation and interests in using the Internet for their assignments and projects (Tsai as
cited in Peng et al, 2006). Therefore, this study will address the following research question:
What are the students’ perceptions of web-based projects at the EMUEPS?


4. Importance of the Research

Using web-based projects is an area which needs further investigation as it takes
considerable amount of students’ and teachers’ time that can be devoted to other instructional
goals and objectives. These web-based projects constitute a part of student portfolio, meaning that
that they have an effect on students overall success at the EPS. Both the administration and the
teachers will benefit from this research as the results obtained from it will help them gain further
insights on the web-based projects and take necessary actions accordingly.


5. Literature Review


5.1. The Use of the Internet in Education

Today, almost all students in all educational settings have a certain experience in using the
Internet for academic purposes. According to Peng et al. (2006), the use of the Internet may affect
students’ learning outcomes in learning environment. The Internet enables students to reach the
recent information in a short time. It also provides students with instant access to an enormous
amount of information and thus it enhances their curiosity and desire to learn more (Yumuk, 2002).


5.2. Students’ Perceptions of the Internet

Previous research studies suggested that students’ attitudes towards the Internet directly
affect their motivation and interests in Internet-based learning. The students may have different
perceptions of the Internet, and these perceptions tend to shape their attitudes and their online
behaviors as well (Johnson and Johnson, 2006).
According to a study done in Sheffield University on students’ perceptions of the Internet
and its use, the most significant findings were related to gender differences. This study revealed
that, female students were unable to find their way around the Internet effectively, thus they often
got lost and felt not in control of what they were doing. (D’Esposito and Gardner, 2000). Hong,
Ridzuan and Kuek (2003) found that, students who have better computer skills in using the
The 3
rd
International Conference on Virtual Learning, ICVL 2008

353
Internet and who perceive Internet as a supportive tool for their studies, have better attitudes
towards using the Internet to improve themselves in their academic studies.


5.3. Internet – based Projects

According to Shiveley and VanFossen (2005), using the Internet in education can serve
different purposes. Internet 1. can be used to access to information, 2. enables students to use
critical thinking skills while using it, 3. can help facilitate collaboration and communication both
within the class and around the world, 4. increases availability to diverse resources and different
perspectives and thus lead to more challenging research projects. 5. can help students to construct
meaning for themselves.


6. Research Methodology

6.1. Identification of the Population

The population under investigation included students who were enrolled in different levels
of the EMUEPS during fall 2007-2008.


6.2. Sample

The sample was selected randomly from class roasters of 2200 students The participants of
this study were 60 students studying at various levels at the EPS.


6.3. Data Collection

Two sets of data were collected for this research study: Internet use survey and student
interviews.


7. Data Analysis and Presentation of Findings

7.1. Internet Use Survey Findings

The main purpose of this study was to investigate students’ perceptions of web-based
projects based on gender, their English level, computer literacy level, having access to a computer
at home, having internet access at home, the aspects of the Internet they usually use and the
frequency of using the Internet. The data collected from the Internet use survey were analyzed
quantitatively through using Independent T-Test and ANOVA on the SPSS program.
The quantitative data examined demographic data and frequencies for all the items in the
survey Demographic Data
The first seven items of this survey were designed to collect “Personal Data”, including
gender, English level, computer literacy level, having access to a computer at home, having
internet access at home, the aspects of the Internet they usually use and the frequency of using the
Internet. An analysis of the characteristics of the target population for the study indicated that 61.7
University of Bucharest and Ovidius University of Constanta


354
% (37) male and 38.3 % (23) female responded to the questionnaire. According to the results of
the descriptive statistics, 16.7 % (10) of the students were enrolled in elementary level, 35 % (21)
in pre-intermediate level, 23.3% in intermediate level and 25% (15) in upper-intermediate level. In
terms of computer literacy level, the results indicate that only 5% (3) of the students identified
themselves as beginner level, 55% (33) as intermediate level and 40 % (24) as advanced level. The
data also reveals that, 83.3 % (50) of the students either have or can access to a computer at home
and 75% (45) of the students have Internet access at home.
In terms of the aspects of the Internet they usually use, using e-mails was marked 36 times,
accessing websites 17 times, search engines 40 times, downloading programs 32 times, playing
audio or video 25 times, chat rooms only 6 times. According to the results, the search engines is
ranked to be the first, e-mail is the second and downloading programs is the third widely used
aspects of the Internet. For the frequency of using the Internet, 63.3% (38) of the students stated
that, they use the Internet daily, 26.7% (16), 1-3 times a week and 10 % (6) stated that they use the
Internet a few times a month.


Frequencies of Individual Items

According to the frequencies of individual items, it is seen that the students who participated
in this study were strongly agree, agree, unsure, disagree and strongly disagree with the survey
items. The frequencies and the percentages of individual items are shown on Table 1 below.

Table 1
Frequencies and Percentages of individual Items


Strongly
Agree
Agree Unsure Disagree
Strongly
disagree
ƒ % ƒ % ƒ % ƒ % ƒ %
1. I feel very confident
of my abilities to use
the Internet for my
projects.
1
7
28.3 36 60 5 8.3 1 1.7 1 1.7
2. Using the internet
for my projects is time
consuming
7 11.7 32 53.3 9 15 5 8.3 7 11.7
3. I can easily access to
any kind of
information when I use
the internet
3
0
50 24 40 5.1 8.3 1 1.7 – –
4. I prefer using other
sources for my projects
than the internet
5 8.3 16 26.7 8 13.3 21 35 10 16.7
5. I get very nervous
when I use the internet
3 5 7 11.7 10 16.7 23 38.3 17
28.3

6. It is fun to use the
internet for my
projects
1
4
23.3 30 50 10 16.7 3 5 3 5
7. Finding appropriate
information for my
project on the internet
is difficult
2 3.3 9 15 17 28.3 21 35 11 18.3
The 3
rd
International Conference on Virtual Learning, ICVL 2008

355
8. There is no need to
use printed materials
when you have the
internet for your
projects
9 15 14 23.3 13 21.7 19 31.7 5 8.3
9. I always have
problems with
computers when I use
the internet
4 6.7 8 13.3 8 13.3 23 38.3 17 28.3
10. I don’t know how
to make best use of
search engines for my
projects
4 6.7 12 20 11 18.3 26 43.3 7 11.7
11. Using the internet
helps me improve my
English
2
5
41.7 26 43.3 4 6.7 1 1.7 4 6.7
12. I don’t like using
the internet for my
projects because there
are lots of unknown
words
4 6.7 7 11.7 8 13.3 20 33.3 21 35

The data gathered from individual items indicate that the 88.3 % of the students feel
confident of their abilities to use the Internet for their projects, 65 % of the students agreed that
using the Internet for the projects is time consuming, 90 % stated that they can easily access to any
kind of information when they use the Internet. 66.6 % of the students disagreed with the idea that
they feel very nervous when they use the Internet, 73.3 % said that it is fun to use the Internet for
their projects. 53.3 % stated that finding appropriate information for their project on the Internet is
not difficult for them. 38.3 % of the students said that there is no need to use printed materials
when you have the Internet. On the other hand, 40 % of the students disagreed with this item. 66.6
% of the students said that they do not have problems with computers when they use the Internet.
Only 20 % stated that they have problems with the computers. 55 % disagree with item 10 which
is ‘I don’t know how to make best use of search engines for my projects’. Only 26.7 % of the
students agreed with this item. For item 11, nearly all, 85 % of the students stated that using the
Internet helps them improve their English. For the last item, only 18.4 % of the students stated that
they don’t like using the Internet for their projects because of so many unknown words and 68.3%
stated that they disagreed with this idea.
In this research, the independent t-test and ANOVA were used in order to be able to analyze
the differences between dependent and independent variables.


T-test of Individual Items

The results of the t-test indicates that there is no significant difference between male and
female students in the way they perceive the use of Internet for their projects as all obtained values
are higher than the standard value which is 0.05
The results of the independent variable which is ‘having computer at home’, reveals that
there is no significant difference between the dependent statements and ‘having computer at home
except the values of ‘I feel very confident of my abilities to use the internet for my projects’ (.040)
and ‘I don’t like using the Internet for my projects because there are lots of unknown words’ (.027)
because all the other obtained values are higher than the standard value: 0.05.
University of Bucharest and Ovidius University of Constanta


356
For question 5 which is related to ‘having Internet access at home’, the results indicate that
there is no significant difference between this item and the dependent variables except for the three
items which are ‘I feel very confident of my abilities to use the Internet for my projects’ (.012), ‘I
don’t know how to make best use of search engines for my projects’ (.034), ‘Using the Internet
helps me improve my English’ (.040) as all these three values are below the standard value: 0.05.


ANOVA of Individual Items

According to ANOVA results which was done for the English level of the students, there is
significant difference between students’ English level and the dependent items 2 ‘using the internet
for my projects is time consuming’ and 3 ‘I can easily access to any kind of information when I
use the internet’ as both values are .038 which is smaller than the standard value 0.05. In terms of
computer literacy level, all the values are above the standard value 0.05, except item 9 ‘I always
have problems with computers when I use the Internet’ because the obtained value .000 is far
below the standard value 0.05. Thus, we can say that there is significant difference between the
students’ computer literacy level and item 9. The results of using different aspects of the Internet
reveal no significant difference as all the obtained values are higher than the standard value, 0.05.
For the last independent item which is the frequency of using the Internet, the results of the
ANOVA indicate that there is meaningful difference only for items 8 (.050) ‘There is no need to
use printed materials when you have the internet for your projects’ and 9 (.024) ‘I always have
problems with computers when I use the internet’. The first value is equal to the standard value
and the second one is below it.


7.2. Student Interviews

The data collected from the interviews were analyzed based on the theme being
investigated. The data gathered from the interviews reveal that, all students have used the Internet
to complete at least three projects on various topics. When approaching research on the Internet,
they used search engines mainly ‘Google’. The data gathered from the interviews can be grouped
under three headings:


a. Students’ feelings towards the Internet

The results of the Internet indicate that most of the students were happy with using the
Internet for their portfolio projects. 48 students out of 60 stated that using the Internet is more
enjoyable than using books. 55 students out of 60 stated that using the Internet enables them to
practice reading and learn new words. According to the data gathered, the students who are
enrolled in higher levels i.e. intermediate and upper-intermediate have less difficulty in searching
the web for their projects. 8 students out of 10 elementary students said that they cannot
understand the materials on the web because of their low level of English. Some of them even
confessed that they access to Turkish sites, find appropriate materials and then try to translate them
into English. Some of the student responses were as follows:
Student 1: “…I don’t understand anything from these web pages I prefer to use books than
the Internet.”
Student 12: “…All my resources are in Turkish and now I need someone to translate them
into English”
The 3
rd
International Conference on Virtual Learning, ICVL 2008

357
b. Difference between male and female students

According to the data collected from the interviews, the female students seemed to have
more difficulties in using the computers especially when they encounter with some technical
problems.
Some female students responded in the following way:
Student 23 : “…my boyfriend always helps me because he is good at using the computer”
Student 39: “… whenever I access to the Internet, something happens and the computer
stops working.”
On the other hand, the male students seemed to be more confident in using a computer and
also using the Internet.


c. Having experience in using a computer and the Internet

Also, the students who had previous knowledge and experience on the use of computers and
the Internet seemed to have the self-confidence and the ability to navigate around the web to reach
the relevant information they need. On the other hand, the ones who have recently got acquainted
with the computers and the Internet had difficulty in using the search engines effectively. Only 7
out of 60 students said that they had not had any experience on the use of the computers and the
Internet before they came to study at the EPS.
Some of the student responses are as follows:
Student 4: “…I hadn’t known how to use a computer before I came here, we didn’t have
computers in our high school in Van”
Student 23: “…I don’t know how to use a computer but my friends are trying to teach me”


8. Conclusion and Recommendations

The data collected from the interviews complement the data collected from the survey. The
results gathered both from the interviews and the survey reveal a positive attitude towards the use
of the Internet for portfolio projects. The data clearly indicate that the computers and the Internet
play an important role in their daily and academic lives of the students as most of them have
access to a computer and the Internet at home. The results also reveal that the students can use
computers and the Internet effectively and efficiently. Most students participated in the study
identified themselves as intermediate level in terms of using a computer. Moreover, only a few
said that they have problems in using the Internet. The interview results indicate that Elementary
level students have problems in understanding the sources and materials because of their low level
of English. Some of the Elementary students said that they use Turkish sources from the Internet
and try to translate them into English. This is an issue that has to be considered by the teachers and
the administration at the EPS. One solution of this problem could be to refer low level students to
certain websites which are specially designed for English language learners.
The students participated in this study have positive perceptions of the usefulness of the
Internet. Since students have very positive perceptions of using the web for their projects, more
Internet-based assignments and tasks could be inserted into the EPS curriculum. Previous research
studies suggested that students’ attitudes towards the Internet directly affect their motivation and
interests in Internet-based learning. The students at the EPS seem to be ready for the Internet-
based instruction.
University of Bucharest and Ovidius University of Constanta


358
The results of the statistical tests which are frequencies, independent t-test and ANOVA
indicate no significant difference between female and male students in the way they perceive the
Internet. The results reveal no meaningful difference between the most statements and the
questions: “do you access to a computer at home?” and “do you have Internet access at home?”
Only a few values show significant difference between the statements and the two questions. The
results of the ANOVA reveal no significant difference between English level of the students, their
computer use levels, different aspects of the Internet and frequency of using the Internet and the
statements. However, there are some values which are below the standard value 0.05.



REFERENCES


D’ESPOSITO, J. & GARDNER, R. (1999), University students’ perceptions of the Internet: An
Exploratory study, The Journal of Academic Librarianship 25(6), 456-461, Retrieved
December 27, 2007 from ScienceDirect database.
HONG, K-S., RIDZUAN, A. A. & KUEK, M-K. (2003), Students’ attitudes toward the use of the
Internet for learning: a study at a university in Malaysia [Electronic version], Educational
Technology and Society, 6(2), 45-49.
JOHNSON, G. E. & JOHNSON, J. A. (2006), Personality, Internet experience and e-
communication preferences, paper presented at the Annual Conference at the International
Association for Development of the Information Society (ERIC Document Reproduction
Service No. ED494002).
PENG, H., TSAI, C. & WU, Y. (2006), University students’ self-efficacy and their attitudes
toward the Internet: the role of students perceptions of the Internet [Electronic version],
Educational Studies, 32(1), 73-86.
SMALDINO, S. E., RUSSELL, J. D., HEINICH, R & MOLENDA, M. (2005), Instructional
Technology and Media for Learning. Pearson Merrill Prentice Hall, New Jersey.
VAN FOSSEN, P. J & SHIVELEY, J. M. (2005), Towards assessing Internet use in the social
studies classroom: developing an inventory based on a review of relevant literature,
Contemporary Issues in Technology and Teacher Education Journal (ERIC Document
Reproduction Service No. ED490637).
YUMUK, A. (2002), Letting go of control to the learners: the role of the Internet in promoting a
more autonomous view of learning in an academic translation course, Educational
Research, 44(2), 141-156. Retrieved December 21, 2007, from Routledge database.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

359
Benefits of Using Self-Study Centres on Language Learning

Zehra Unveren, Gülen Onurkan Aliusta, Fatma Basri

Eastern Mediterranean University, English Preparatory School, North Cyprus
E-mail: zehra.unveren@emu.edu.tr


Abstract
Self-study centres play an important role in language learning. In these centres,
students have the opportunity to practice the target language through a wide variety
of software on the Internet. Since classroom time is limited, students do not have
much time to use and practice the target language adequately in class. Therefore,
self-study centres enable students to study at their own pace and according to their
individual needs outside the class. In the English Preparatory School of Eastern
Mediterranean University, it has been observed that some students continually make
use of the available resources and technology offered in these centres while others
are not aware of the benefits. The aim of this research was to investigate the
relationship between students’ achievement and the amount of time spent in these
centres. The participants of this study were 50 students studying at the Intermediate
level of the EPS program. For this research study, it was necessary to collect two
sets of data: student interviews and test results. The results collected from the data
clearly indicated that there is a positive correlation between student test results and
the amount of time spent in the centres. According to the data, the students who
regularly use the self-study centres performed better in the tests.

Keywords: Self-study centres, Independent learning, Autonomy.


1. Introduction

Recently, self-access centres in higher education institutions have become popular
throughout the world. In these centres, students are expected to supplement limited class contact
time with independent study through the use of technology such as satellite television, Internet
access, CALL materials, on-line dictionaries etc. Students in higher institutions usually receive 3
to 5 hours of taught classes a day and are expected to work in these centres in order to be able to
extend and enhance what they have been taught in class. Self-access centres give students the
opportunity to study at their own pace according to their weaknesses and needs. As Carbone
(2000) puts it, these centres enable students to develop the capacity to make decisions, reflect,
manage and extend their learning beyond the classroom. However, although self-access centres are
well-resourced in terms of educational technology and materials, they are generally under-used
(Souto and Turner, 2000).


1.1. Teaching/learning environment at the EPS

Eastern Mediterranean University (EMU) is an English medium university in North Cyprus.
Students who are not proficient in English are required to study at the English Preparatory School
University of Bucharest and Ovidius University of Constanta


360
(EPS) in order to be able to study in their chosen departments. Therefore, the EPS aims to equip
students with necessary skills and strategies which will help them survive in their departments.
In the English Preparatory School of the Eastern Mediterranean University (EMUEPS) in
North Cyprus, there are two Student Self-Study Centres which are equipped with lots of
educational materials that enable students to study English outside the classrooms. Students can
have access to these centres any time between 8.00 and 17.00 during weekdays. In these centres,
there are full-time English language teachers who are always available to guide and help students
with the activities/materials they want to work on.
These centres are comprised of a multi-media section, a library, a listening section, a TV
room and a speaking section. Multi-media sections are the most popular and most widely used
sections by students. In the multi-media section, students are provided with a good selection of
software such as Moodle to assist them in practicing grammar, vocabulary, reading, writing and
listening. Other software such as encyclopedias, dictionaries and documentaries are also available.
The main aim of these centres is to encourage students to take responsibility for their
learning and lead to learner autonomy which is an essential requirement of higher institution.
According to Chia (2005) autonomy is an important educational goal and there is a close
connection between learner autonomy and effective learning.


2. Statement of the Problem

The majority of Turkish students undergo the process of learning through traditional
educational methods (Yumuk, 2002). Most of the students at the EPS are mainly Turkish who are
from the traditional educational background where students sit in rows and teacher is the sole
authority in class. They think teacher is the only source of information and they expect to learn
everything from the teacher. “The teacher-student relationship is mainly limited to one-way
channels of communication in which teachers transfer information to learners” (Yumuk, 2002:
143). Moreover, students are not aware of their weaknesses in language learning and most lack the
necessary learning skills and strategies required for them to be autonomous and self-independent
learners in the learning process.
Majority of the students at the EPS are from traditional educational background and are
often far from being autonomous. They have teacher dependent learning habits and thus they
expect to learn everything during the class hours from the teacher. They are not aware of the
benefits of working independently and adopting the strategies that will help them to become
autonomous learners. Because of the students’ educational background, it has been observed that
while some students in the EPS continually make use of the self-study centres, most do not seem
to be aware of their benefits.


3. Aim of the Research

The main aim of this research study is to investigate whether there is a correlation between
student progress in language learning and their regular use of these centres. This study will address
the following research question:
Is there a correlation between student progress in language learning and their regular use of
the Self-Study Centres?


The 3
rd
International Conference on Virtual Learning, ICVL 2008

361
4. Importance of the Research

The results of this study play an important role in foreign language learning as it will inform
both students and teachers about the impact of independent learning through the use of self-study
centres on language achievement. Also, this is an area which needs further investigation in order
to be able to encourage and train students to use these centres more effectively. It is discouraging
for the teachers and the administration to observe that, although there are two fully equipped
centres in the EPS, students do not seem to make use of these centres for academic purposes. The
results of this study will be beneficial both for the administration and teachers, and the students in
the EPS as it will help students gain further insights into becoming autonomous learners through
the use of these centres regularly, and thus, it will enable the EPS administration to take necessary
actions accordingly.


5. Literature Review

5.1. The Importance of Self-Study Centres in Language Learning

“Self-access centres (SACs) are playing an increasingly pivotal role in supporting the (self-)
study of languages” (Reinders and Lewis, 2006: 272). Kell and Newton suggested that the main
aim of the Self-Study Centres is to provide “pathways” in order to guide learners on how to use
these centres.(1997: 48). “… the pathways encouraged students to develop the skills needed to
organize their own learning and would help move them on to more autonomous learning methods”
(Kell and Newton, 1997:52). Self-Study Centers enable students to take responsibility for all the
decisions regarding language learning such as determining the objectives, selecting materials and
activities, deciding on the place, time and pace of learning (Chia, 2005).


5.2. Use of Technology in Self-Study Centres

In the Self- Study Centres, use of computers, mainly for its software and the Internet, plays
an important role in language learning (Chia, 2005). “Inspired by rapid development of technology
from the 1980s, computer has now become an influential component of second language learning
pedagogy” (Lai and Kritsonis, 2006: 1). Lai and Kritsonis (2006) pointed out that computers
enable second language learners to develop their linguistic skills, affect their attitude towards
language learning and also build their self-instruction strategies and self-confidence through the
use of various communicative and interactive activities. As Carbollo-Calero (2001) mentioned,
language teachers should consider the use of computers as an important supplementary tool for
learning and bear in mind the role of the teacher as a facilitator rather than the only source of
information (as cited in Ayres, 2002). Lai and Kritsonis (2006) further stated that it is important to
explain the advantages of computer technology to teachers and students because without giving the
necessary guidance it would be difficult for teachers and students to become aware of the benefits
of computer technology for language learning.


5.3. The Reasons for not Using the Self-Study Centres

It has been observed that most students in the EPS do not seem to use self-study centres regularly
because of various reasons. According to Souto and Turner (2000) the following are the main reasons:
• working independently is unfamiliar; students need time to adapt;
University of Bucharest and Ovidius University of Constanta


362
• working alone is unnatural in language learning; the human and social dimension is
missing, as are the immediate feedback and encouragement from the teacher;
• students need preparation and training to become independent learners;
• students need convincing that learning take place without the teacher;
• students need training in study skills (p. 388).


6. Research Methodology

6.1. Identification of the Population

The population under investigation included students who were enrolled in the intermediate
level of the EPS during spring 2007-2008 academic year in Eastern Mediterranean University in
North Cyprus.


6.2. Sample

The sample was selected by the method of random sampling from class roasters of students
studying at the intermediate level at the EPS. The participants of this study were 50 students who
are currently enrolled in the intermediate level of the EPS program.


6.3. Data Collection

Two sets of data were collected for this research study: student interviews and students’ test
results. The researchers decided to use more than one data collection method in order to be able to
reach more accurate, valid and reliable data.


Student Interviews

The main aim of the student interviews was to find out the frequency and purposes of
students using these centres. The students were asked the amount of time they spent in the
centres and what they used this time for.


Test Results

In the final exams, the students were tested on four language skills plus language features
and vocabulary, and the test scores that the students obtained from their final exams were used in
this study.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

363
7. Data Analysis and Presentation of Findings

The main purpose of this study was to investigate the correlation between student progress
in language learning and the frequency of their use of the Self-Study Centres. For this study, it was
necessary to gather two types of data: test scores and student interviews.


7.1. Test Scores

The data collected from the student test results were analyzed quantitatively through using
correlation coefficient on the SPSS program. According to the result obtained from the Pearson
product-moment correlation coefficient (r = 634.), there is positive correlation between students’
test results and the frequency of their use of these centres. The scatter diagram was also prepared
to determine the degree of correlation between the two variables. The scatter diagram also shows
that there is moderate positive correlation. The results of the correlation between the two variables:
test scores and frequencies and the scatter diagram, are displayed on table 1 and table 2.

Table 1
The Result of the Pearson Product-Moment Correlation Coefficient

test scores frequency
test scores Pearson Correlation 1 .634
Sig. (2-tailed) . .000
N 50 50
frequency Pearson Correlation .634 1
Sig. (2-tailed) .000 .
N 50 50
** Correlation is significant at the 0.01 level (2-tailed).

Table 2
The scatter Diagram
frequency
6 5 4 3 2 1 0 -1
t
e
s
t

s
c
o
r
e
s
70
60
50
40
30
20
10

University of Bucharest and Ovidius University of Constanta


364
7.2. Student Interviews

All 50 participants were interviewed and they were asked the following three questions:
1. How often do you use Self-Study Centres?
2. How many hours do you usually spend there?
3. What do you use these centres for? Why?
Later, the answers were analyzed based on the theme being investigated.


a. Frequency of Students’ Using the Self-Study Centres:

According to the results obtained from student answers, out of 50 students, 21 students
stated that they visit these centres 3 or more times a week for academic purposes. 25 of the
students said that, they occasionally visit the centres and their main aim is to check their mails or
learn the exam results on student portals. Only 4 students stated that they never go and study in
these centres.


b. The Amount of Time Students Spend in the Self-Study Centres

Students who make use of these centres said that they spend there about one hour each time
they visit these places.


c. Students’ Aim for Using Self-Study Centres

The results of the student interviews reveal that only 31 students out of 50 use these centres
for academic purposes. They mainly use educational technology offered in these centres: satellite
TV, online dictionaries, CALL, Internet, etc. The most commonly used ones are the software and
the Internet. Software such as Moodle and Eagle (an inhouse program) are the most popular ones.
These programs enable students to work independently at their own pace for the purpose of
improving their weak areas. The Internet is also used by the students for a variety of reasons such
as doing research for their projects, reading online newspapers, chatting and for entertainment.
Exam practice is another area which receives lots of attention from students usually during exam week.
Students do exam practice such as listening, reading and vocabulary through using related websites.


8. Findings and Conclusion

The data gathered from both test results and student interviews indicate that regular use of
these centres has a positive effect on student progress in English language learning. When the test
results and the interviews are compared, it is clearly seen that the students who had regularly visited
these centres for academic purposes outperformed others who had only visited these places occasionally.
It was also noticed that the students who use these centers once or twice a week do not
seem to use these centers effectively, and thus do not achieve their purposes which demotivates
them to spend more time in these places. The reasons pointed out by the students for not using
these centres regularly and effectively are as follows:
1. As they are teacher dependent students, they are not used to studying independently. They
prefer to be directed by a teacher in their studies and receive feedback for the outcome of their study.
The 3
rd
International Conference on Virtual Learning, ICVL 2008

365
2. They do not know what resources are available in these centres.
3. They are not happy with the atmosphere. The centres are not very appealing to them.
4. They lack the necessary orientation and training for the centres.
5. They do not have the necessary computer skills.
6. They prefer working in collaboration with their peers.
In fact, all these reasons arise from the fact that a considerable amount of students in the
EPS are not aware of the benefits of using the self-study centres in language learning.


9. Suggestions

9.1. Orientation on the use of the resources in the centres

At the beginning of each academic year, students should receive orientation on the use of
educational resources available in these centers. It is important to give students hands on tasks to
familiarize them with these resources rather than giving pure theoretical information.


9.2. Training

Training should be given in two different areas: 1. Training on the use of educational
technology such as using computers and the Internet. 2. Training on the benefits of using self-
study centers. Students should be informed about the importance of being autonomous in language
learning. Classroom time is limited so students should know that they need to get exposed to the
target language as much as possible if they want to be self-sufficient in their academic studies.
They should be able to identify their weaknesses and use the resources in these centers to do
extra/remedial practice outside formal language learning environment.


9.3. Allocated class hour

In order to give the required training and to encourage the students to use these centers
effectively and regularly, each class should be allocated one class hour a week to spend with their
teachers in these centers. This will enable students to get used to studying in these places and
realize the advantages of using these centers



REFERENCES


CARBONE, A. (2000), Transforming students into self-directed, independent adult learners,
retrieved April 21, 2008, from http://www.ala.asn.au/commentaries/Carbone0111.pdf.
CARRIER, M. (1997), ELT online: the rise of the Internet, ELT Journal 51,3, 279-301.
CHIA, C. S. C. (2005), Promoting independent learning through language learning and the use of
IT, Educational Media International42, 4, 317-332.
KELL, J. & NEWTON, C. (1997), Roles of pathways in self-access centres, ELT Journal 51, 1, 48-53.
LAI, C. & KRITSONIS, W., A. (2006), The Advantages and disadvantages of computer technology in
second language acquisition, retrieved May 1, 2008, from http://www.nationalforum.com/
University of Bucharest and Ovidius University of Constanta


366
Electronic%20Journal%20Volumes/Cheng-Chieh%20Lai%20The%20Advantages%20and%20
Disadvantages%20of% 20Computer%20Technology.pdf.
REINDERS, H. & LEWIS, M. (2006), An evaluative checklist for self-access materials, ELT
Journal 60, 3, 272-278.
SOUTO, C. & TURNER, K. (2000), The development of independent study and modern languages
learning in non-specialist degree courses: a case study, Journal of Further and Higher
Education 24, 3, 385-395.
YUMUK, A. (2002), Letting go of control to the learners: the role of the Internet in promoting a more
autonomous view of learning in an academic translation course, Educational Research 44,
2, 141-156.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

367
Web Based Simulator for Virtual Company-Market Game

Idehara, Norimichi
1

(1) Department of Management and Information Sciences
Tama University
4-1-1, Hijirigaoka, Tama, Tokyo 206-0022, JAPAN
E-mail: idehara@tama.ac.jp


Abstract
In this paper I describe our virtual company management game simulator. The
simulator is used as a tool for lecturing basic economy and accounting in our
university. The system is characterized by two points. One is that the participants
are public customers as well as company workers, and their behavior construct the
market. Researching behavior of others is required to determine the company
strategy. The game will not result in a predefined success / failure, but in more
dynamic market situation; sometimes such as inflation / deflation. The other point is
that each participant defines his/her own preference vector at the beginning of the
game, and each product has its feature vector. These vectors introduce the market
positioning. With this system, students can learn (1) the relationship between
consumers, companies and market, (2) the importance of company strategy, (3) to
understand financial statements, and (4) the management of team. The system is
based on a multi-user web database system. Each participants is given the ID and
the password, with which we can track the participant's behavior. The participants
can change the price of their product and purchase products at anytime, anywhere
with a web browser.

Keywords: company simulator, market simulator, web based learning.


1. Introduction

Tama University Business Game (TamaBG) is a tool for lecturing basic economy and
accounting to freshmen in our university. The participants is a company worker in a virtual world
with monetary unit Tim(T). The system has been used for nine years [Saito], and each year about
200 to 300 students take the program. The program is a combination of lecturing basic theories and
practice in the simulation game. In this paper, the unique feature of this simulation system is described.
Before introducing TamaBG, most freshmen suffered from lacking in the understanding
what market is and how companies are really working. This system aims to enable the participants
to reach practical understanding in market mechanism, company strategy, financial statements and
team management.
There are many company simulators for lecturing company management or economy. Most
of them have a well-defined external market [Kobayashi]. Because our system aims to make the
participants to have more practical understanding in the dynamic relationship between market and
consumers, such market model is not suitable. In TamaBG, the participants are consumers as well
as company workers. The compilation of participants' decision to purchase a product is the market.
University of Bucharest and Ovidius University of Constanta


368
Virtual Enterprise Australia (VAE) is a similar system to TamaBG, such that participants can play
both roles, though VAE focuses mainly on improving the abilities as business workers. There is no
reward to take part in the simulation as a consumer. In TamaBG, the participants as consumers are
required to maximize their benefit point (Bp) that is gained through the purchase of products.
Each participant defines preference vector at the beginning of a game. Companies
simultaneously design their product position as the product feature vector. Bp is gained best when
preference vector and product feature vector are parallel. Thus, the benefit for a product varies
from players to players, which leads students to realize the importance of market research and
product positioning strategy.
We implemented the system with a database server and a web interface server. The participants
can access to the simulation system at anytime from anywhere to purchase products. The price of
the product can be changed as well, so dynamic pricing strategy is important for company.


2. Program Schedule

The standard game of TamaBG consists of eight terms each of which lasts for a week. At
the beginning of a term, participants have lectures in two periods: three hours. In the first period, a
lecture for the topic of the week is given. In the second period they access to the simulator as
company workers to find out the result of the previous term and start discussing on the strategy for
the new term. After the discussion and producing a new product, they start their role of consumers.
They spend the rest of term purchasing and checking the price of competing products with their
web browser. In our university, every students are supplied their own notebook computers. The
simulator will stop on the last day of the term to commit the term rotation procedures.
At the beginning of the program they have an experimental game for two terms. At the end
of program, the stockholders' meeting to which all the participants must attend is held and each
company must explain their strategy (Figure 1).





Figure 1. Program Schedule
The 3
rd
International Conference on Virtual Learning, ICVL 2008

369
3. Consumers Role

A consumer aims to maximize his/her benefit point (Bp). Through the process, they are
lead to understand the relationship between consumer and market, the effect of economical decline
and the role of investment.
At the beginning of the game, a consumer determines his/her preference vector. Purchasing
a 'good' product at low price is the best way to gain Bp, though the definition of 'good' varies with
the preference vector.
The salary for the consumer is paid automatically from the company account. The company
benefit may be distributed as a bonus. To gain the bonus, a participant must seek for more benefit
for their company.
A consumer may also invest to a company and expect the benefit distribution as well. Self
investment to their own company is not prohibited, and in many cases it is a good choice. Because
the system can be accessed through web interface, consumers can, and encouraged to, check and
purchase products at anytime from anywhere. Most students tend to make the choice at the first
two days of the term, though. (Figure 2. Note that a term begins on Wednesday).
The rank of Bp is continuously updated on a web page, so that students are highly
motivated to gain the point.




4. Company Workers Role

A company consists of 4 to 7 participants. Discussing about the financial and marketing
strategy enables the students to understand practical decision making process, financial report, market
analysis, product positioning, pricing strategy, advertisement, product life cycle and compliance.
Each company worker should be one of the positions: president, product manager, financial
manager, and sales manager. Different assignments are placed to each position.
At the company foundation, a product category must be chosen from eight categories. In
2006 they are: house, fine art, car, furniture, computer, leisure, clothing and book. The category
determines the base unit cost and the maximum number of production, though they are tuned to be
simultaneous condition. A company may change its category by disposing half of its product
equipment and buy a new product line. Excessive competition is inevitably observed in some
Figure 2. Sales Ratio (%) in Each Category
University of Bucharest and Ovidius University of Constanta


370
categories. Furthermore, the consumers have maximum affordable unit count for each product
category. Therefore a company should carefully analyse the maturity of the market.
At the beginning of each term, the workers of each company have a meeting to decide the
strategy. Each term they can produce only one product. At the meeting they have to decide:
Benefit distribution
Determine sales and general administrative (SGA) expense: research and development(R&D),
advertisement, and product line improvement
Product positioning
SGA has sigmoid improvement effects on the three base parameters: unit production cost,
maximum productive count and satisfaction (Figure 3) Each expense has minimum value of 70
Tim(T) and maximum value of 210T. Expense for R&D and product line improvements can be
applied to either of two parameters, so there are many alternatives to compare (Table 1).
Product positioning is determined by setting the product feature vector. There is no cost for
the positioning. The length of the vector is normalized.



Some students try to gain Bp by suddenly reducing the price of their own product,
purchasing them and restoring the price. Some students try to give themselves the bonus more than
the benefit or under the passive balance. Those actions which are semi-automatically inspected by
the system result in breach of trust (BOT) penalty to the student and are announced.

Table 1
Improvement Effects of SGA

Unit Production Cost Max. Productive Count Satisfaction
Advertisement O
R&D O O
Product Line O O




Figure 3. Satisfaction Improvement Factor to Ad Expense
The 3
rd
International Conference on Virtual Learning, ICVL 2008

371
5. Preference / Feature System

Most unique feature of TamaBG is the preference/feature system. The system is introduced
to TamaBG in 2007. Earlier games resulted in a speed race to find the potential market category
often with some luck, to produce “best” satisfaction product because of diminishing improvement
effect, and to win. Such unfavourable situation is improved by this system modification.
Each participant defines his/her own three dimensional preference vector (P) at the
beginning of a game. P is fixed throughout the game. The element of the vector represents
“functionality”, “design” and “celebrity”. Companies simultaneously design their product position
in these three features as the product feature vector (F). Bp that is gained by a consumer with
preference vector P with a product of satisfaction value S and product feature vector F is as follows:

) ( 3 F P S B
p
• =

The constraint of P and F are different. Note that the constant length of F justifies its
costless determination by the company, while the length of P varies:

( )
( ) 1 1 1 00 . 1 ) (
57 . 0 ) (
1
1
= = = =
= = =
=
= + +
c d f
c d f
c d f
P or P or P P Max
P P P P Min
F
P P P


A consumer must consider to have whether general interest or special interest. Moderate Bp
is gained by any products with general interest, while with special interest some products give high
Bp and some gives nearly no Bp.

Bp is gained best when P and F are parallel. Thus, the benefit for a product varies from
players to players. Consumers must consider not just price and satisfaction value of a product, but
whether the product matches their P as well.


6. Software / Hardware Implementation

Table 2
Sample of Preference/Feature Factor

Products Consumers Bp
Product Fc Ff Fd S A ( .3, .3, .4 ) B (1, 0, 0) C (0, 0, 1)
A 0.6 0.6 0.6 100 90 100 100
B 0.1 1 0.3 100 71 24 47
C 1 0.2 0.1 100 69 167 21

The system has been suffering stability problem due to the concentrated access during the
lecture period. After the improvement both on software and hardware, current system is stabilized.
University of Bucharest and Ovidius University of Constanta


372
The system consists of two servers; one is a database server c5.tama.ac.jp with FileMaker
Pro 7 (FMP7) / FileMaker Server 7 Advanced (FMS7A) and the other is a Apache web interface
server iis.edu.tama.ac.jp. The lecture is also supported by a bulletin board system XOOPS.
Originally all the system was running on a server and all program was coded in FileMaker
script. Separating the web service and exporting built-in FileMaker script to PHP on web interface
server improved the stability. The web server accesses FMS7A with PHP library: FX.php.
The database server is still the bottleneck of the system. It gets unstable at the peak access
of 30 requests per second during the lecture period. The response would be close to 30 second per
request at worst, and many students are observed to reload the page. After the frequent reload
actions, the database server is observed to stop responding. Recent setting modification that
prohibits another request from an IP address while processing one request improved the stability
problem. For this purpose, Apache module: mod_limitipconn is introduced.
Further optimization to decrease the query between the web server and the database server
are planned.


7. Next Goal

Next goals are as follows.
Improvement of response: this is done by upgrading the database server to multi-core,
multi-CPU and server software to FileMaker Server 9. The FileMaker script procedures should
exported to the web server.
Introducing incentive for market operation on other days: the operation is now practically
limited to two days and dynamic pricing strategy is not effective.
Introducing incentive for “special interest” preference: the students are observed to have the
tendency to choose “general interest” preference such as (.3, .3, .4). The tendency spoils the
market research experiment. More beneficial tune for “special interest” preference is required.
Creating stock market, so that company can raise its capital: when a company wishes to
raise its capital, there is no way other than self investment and self purchase.



REFERENCES


KOBAYASHI, K. (1992), A Business Game in a New Style, Journal of Commerce 41, 2, 23-44.
SAITO, H. (1999), System for Business Game in Virtual Market: Basic System for TamaUniversity
Business Game, Tama University journal of management and information science, 3, 39-49.
Virtual Enterprise Australia: http://www.anpf.cit.act.edu.au/
FX.php: http://www.iviking.org/FX.php/
FX_charset.php, FX.php Japanese Library: http://msyk.net/fmp/fx_ja/

The 3
rd
International Conference on Virtual Learning, ICVL 2008

373
Virtual Learning Space
with Semantic Web Technologies

Ioana Andreea Stănescu
1
, Antoniu Ştefan
1
, Veronica Ştefan
2


(1) Advanced Technology Systems, 222 Calea Domneasca
Târgovişte, ROMANIA
E-mail: ioana@ats.com.ro
(2) Valahia University of Târgovişte
2 Carol I Street, Târgovişte, ROMANIA


Abstract
The learning process enables us to participate successfully in life, work and relevant
communities. In the last decades, most applications that were developed sustained
mainly the formal learning within educational institutions and training centres. As
the core of practice and the place where we test our knowledge is the workplace
environment, the purpose of this paper is to present the virtual learning space within
a company and how learning can sustain and increase its efficiency. While formal
learning is strongly needed as it sets the foundation landmarks of our education,
informal learning builds practical experience, which translates in skills and abilities
adapted to the workplace environment. Informal learning is the unofficial,
unscheduled way people learn to do their work. In a networked economy, companies
need to understand that learning is the competitive advantage. Learning is a
productive adaptation to change. This represents learning with a purpose, learning
that can extract the earning out of learning. EDU.PROJECT developed by
Advanced Technology Systems, sustains the lifelong learning process by creating
new learning spaces with semantic web technologies. In this paper we shall explore
how the Web evolution can make workplace learning adaptable and flexible and
how it has the potential to increase revenues, cut costs, accelerate innovation and
develop the flexibility of a company.

Keywords: Informal Learning, Semantic Web Technology, Uniform Resource
Identifier, Resource Description Framework.


1. Virtual learning space within a company

Industries increasingly rely on research and innovation. Innovation is characterized by an
intense, collaborative process of generating and exploring ideas meant to contribute to the solution
of particular problems. Innovators go through cycles of divergence, in which new ideas are
generated and explored, and convergence, in which new ideas are valued and detailed. These
cycles are built on knowledge elicitation (formal learning) and knowledge sharing (informal
learning). In this respect, we propose herein to analyze the virtual learning space within an
organisation to identify the strengths of efficient learning modelling.
Education, whether formal or informal, implies complex combinations of interactions
between learners, instructors and technologies. At the present, the Internet can be defined as a
University of Bucharest and Ovidius University of Constanta


374
hard-working provider of information that lacks efficiency because it delivers information it cannot
comprehend. The “Semantic Web”, a term coined by Tim Berners-Lee, refers to a vision of the next
dramatic evolution of web technology where intelligence and meaning is being added to the display and
navigational context of the current World Wide Web. Semantic Web developments can be used to
build attractive and more successful educational infrastructures that facilitate access to content.


1.1. Learning Alternatives within Organisations

People are designed to never stop learning and exploring (Medina, 2008). Learning is what
enables people to participate successfully in life and work. It is a knowledge-age survival skill
(Cross, 2006) and companies have to consider the importance of its sustainable development. Most
learning doesn't occur during formal training programs, but through processes not structured or
sponsored by a school or an employer. To truly differentiate between formal and informal, we also
find it valuable to examine what is learned intentionally or accidentally.
Formal learning includes the hierarchically structured school system that runs from primary
school through the university and organized school-like programs created in business for technical
and professional training.
Informal learning describes a lifelong process whereby individuals acquire attitudes, values,
skills and knowledge from daily experience and the educative influences and resources in his or
her environment, from family and neighbours, from work and play, from the market place, the
library and the mass media.



Figure 1. Learning Alternatives (Conner, 2008)

Intentional learning is the process whereby an individual aims to learn something and goes
about achieving that objective.
Accidental learning happens when in everyday activities an individual learns something that
he or she had not intended or expected.

The 3
rd
International Conference on Virtual Learning, ICVL 2008

375
1.2. The Informal Learning Perspective

Informal education is a lifelong process by which every individual acquires and
accumulates knowledge, skills, attitudes and insights from daily experiences and exposure to
environment. We learn at work, in house, en route, on the run, in context, in situ, through search,
by accident, from children, in press, across TV, by mistake, ah ha! (Conner, 2008). We learn by
reading, talking with experts, talking with peers, email or other written correspondence, and
through a coach or mentor.
Generally, informal education is unorganized, unsystematic and even unintentional at times,
yet accounts for the great amount of any person’s total lifetime learning – including that of a
highly ‘schooled’ person (Coombs and Ahmed, 1974). Within this