PROCEEDINGS OF THE International Forum cum Conference ON Information Technology and Communication at the Dawn of the New Millennium Bangkok, Thailand 1-4 August 2000 Organized by > Asian Institute of Technology © National Science and Technology Development Agency in Association with: Kasetsart University Khon Kaen University Bangkok University Burapha University Chulalongkorn University Saint John’s University King Mongkut's Institute of Technology North Bangkok Campus Chiang Mai University ‘Thammasat University VVVVVVV vv Sponsored by: Siemens AGInternational Forum cum Conference Information Technology and Communication at the Dawn of the New Millennium August 1-4, 2000 The Imperial Queen's Park Hotel, Bangkok, Thailand Sponsored by: Siemens AG Co-sponsors & Supporters: Acer Computers + Computer Associates + Hewlett Packard + IBM “+ Microsoft + Motorola * Nokia “+ Seagate Technology + Sumitomo “ The Nations + The Imperial Hotels Group International Academy in Engineering Development and Scientific and Management Technology (IAEDSMT) is funded by: MR.YOSHIRO TAKASAKI Educated in Waseda University, is the Chairman of Toyo Seikan Kaisha Ltd. He is also the Executive Director of Japan Federation of Employers Associations (Nikkeiren), Auditor of the Federation of Economic Organizations (Keidanren) and the Councilor of Japan Committee of UNICEF. He has been the Chairman of the Board of Trustees of Toyo Institute of Food Technology. A leading businessman, Mr. Takasaki established the Crown Seal Public Company in Thailand which is a Thai-Japanese Joint Venture. He is also the Chairman of the Bangkok Can Manufacturing Co., Ltd. A recipient of the ‘Medal with Blue Ribbon of the Japanese Government, Mr.Takasaki is a Member of the AIT Board of Trustees.PROCEEDINGS OF THE ——S—————EE EES International Forum cum Conference ON Information Technology and Communication at the Dawn of the New Millennium Bangkok, Thailand 1-4 August 2000 Organized by > Asian Institute of Technology © National Science and Technology Development Agency in Association with: Kasetsart University Khon Kaen University Bangkok University Burapha University Chulalongkorn University Saint John's University King Mongkut's Institute of Technology North Bangkok Campus Chiang Mai University ‘Thammasat University VVVVVVWV vv Sponsored by: Siemens AGThese volumes of the Proceedings are dedicated to H.E. MLL. Birabhongse Kasemsri Former Chairman of the AIT Board of Trustees For his most valued and untainted help in the formation of the International Academy and in guiding many of the Activities which are held so far in a successful manner.International Academy in Engineering Development and Scientific and Management Technology (LAEDSMT) The International Academy in Engineering Development and Scientific and Management Technology (IAEDSMT) is established to activate a truly International Participation of Engineers and Scientists in the development of Asia through the Asian Institute of Technology. The Academy aims to fulfill this mission by: assisting engineering education, research, development, and, innovation in developing countries. promoting the awareness of the role of engineering and science in sustainable economic development. + promoting effective relations beoween the industry, government and universities in technological development. % ensuring employment prospects vis-a-vis productivity and technological development. recognizing the excellence in engineering development and scientific technology and their role in urban and rural development. * For forty years, AIT has served as a United Nations Embassy with multi-national Ambassadors (Trustees, Teachers, Staff and Students) for higher education and research fostering interaction between developed and growing countries, Within this short period, AIT has produced over 10,000 Alumni with post-graduate education. As our world is changing rapidly, creativity, innovation and entrepreneurship are the order of the day and international cooperation is mandatory in solving our host of crucial problems for which the answers lie in the advancement of Engineering, Science, Humanities, Management and Technology. It is anticipated that the International Academy will pave the way for intellectuals from all parts of the world to meet on a common ground with the ultimate goal of constantly updating and advancing the human resources and the environmental concerns of Asia and elsewhere. Such an Academy can also help identify crucial focus areas in research and development and continue to re-orient them in accordance with the needs of the countries in the region. Additional activities include the preparation of educational books, monographs, journals and conducting high level workshops, seminars, meetings and conferences. It is increasingly becoming evident that private sector industries and governments must be more closely knitted with Universities in their varied activities. This can only be achieved if the University curricula and research topics are of a problem solving nature while rooted on a sound fundamental foundation. It is hoped that there would be very strong support from the Industries and the Private Sectors for the activities of this International Academy.Message from Professor Jean-Louis Armand President of Asian Institute of Technology Being the only Regional Post-graduate Institute in Asia, the Asian Institute of Technology has always acted as a catalyst in the development of modern technologies in developing countries. 1 am proud that AIT has done an excellent job over the years in conducting post-graduate teaching, research, and professional activities. All these endeavors are impossible without the much needed Industrial Collaboration and contributions. The Giants of the Information Technology are often admired for their contribution to the so-called “New Economy” which is expected to strengthen University Education not only in using IT Techniques but also in conducting collaborative research and development in this fast growing discipline. As AIT enters a new era in which the strengthening of technical education in developing countries has become an important component of the multi-faceted activities that we are engaged in, I am delighted to see that we now have the opportunity to work more and more closely with IT Industries and this will certainly enrich our contributions towards Human Resources development and other urgent development needs in our Society. I am confident that a very productive deliberation will come out of this conference. My sincere appreciation to all those who have contributed to the success of yet another successful activity in AIT in its continuing vision to work for the region’s development. Message from Professor Huynh Ngoc Phien Dean, School of Advanced Technologies As the Dean of the School of Advancement Technologies, I am very pleased to write this brief message on the occasion of the First International Conference on Information Technology. Needless for me to re-echo the importance of this growing field within our School and the Institute as well as in a global perspective. We are fortunate that many of the Academic Programmes within our School were formulated and developed in partnership with leading industrial organizations such as IBM, Nokia, Seagate, Lucent Technologies etc, We have a full spectrum of the Industrial Sector in Bangkok and also in the region and beyond helping us with our regular Forum and other activities. Indeed Siemens AG is part and parcel of AIT at all levels from the Board of Trustees to the students’ graduation prizes. Through this conference, we are further looking ahead to develop our existing ties and to form new partnerships with other IT Giants and the Leading Universities and researchers. My sincere thanks for all those who have contributed to the success of this event.Message from Professor Worsak Kanok-Nukulchai Dean, School of Civil Engineering On behalf of the School of Civil Engineering, I sincerely thank Siemens AG and the other co- sponsors and supporters for their invaluable support in the organization of this conference which recognizes the significant roles and impacts of Information Technology in all aspects of our life including education at all levels. We have witnessed an exponential growth in the dissemination and acquisition of information through Internet. This mode of information exchange will certainly serve the basis for life-long learning in the years to come. It is no longer possible or even necessary to acquire knowledge once and for all during the initial education and then spend the rest of one’s life applying it. With so many changes going on dynamically in every corner of the world, information technology is at the heart of this phenomenon. The current Asian crisis, in fact, gives engineering education institutes like AIT an opportunity to re-invent its wheels to train a new breed of engineers to cope with 5 keywords resulting from these changes. The foremost of them is Information Technology. The rest are International, Integration, Industrial Partnership, and Innovation. In the last century, industrialized nations became rich because of their competitive advantage due to their great efficiency in production. However, in the new millennium, you have already witnessed a change - a country can only leap to prosperity through information technology. My very warm welcome to all the participants and my best wishes for the success of this Conference. Message from Professor Mario T. Tabucanon Provost, Asian Institute of Technology For all of us, in the 21st century, there is no escaping from the clutches of the IT revolution. IT touches and transforms people's lives, without regard of race, creed or religion, including the man on the street. There are a lot of wondrous things that it has created. But, there are also potentially grave consequences if its power is abused. Clearly, it is our task to promote the former, without losing sight of the latter. Hopefully, if we - from the IT business and IT education sectors - join hands together, through this International Forum cum Conference on Information Technology and Communication at the Dawn of the New Millennium, the task of building a better society, through the wide-ranging applications of IT, will be made possible. With deep gratitude and all best wishes to the organizers, sponsors, speakers and participants.Preface The International Academy on Engineering Development and Scientific and Management ‘Technology, established as a new arm under the Asian Institute of Technology, aims to form a network of Leading Universities as well as State and Private Sector Agencies and Research Organizations around the world. ‘The Academy fully recognizes the current trends in the field of. Information Technology (IT), telecommunications and their revolutionary impacts worldwide in shaping and moulding our society in a meaningful manner. We are witnessing an exponential growth and development in the trade of Internet services and software production and their interaction with modern developments in the telecommunications industry. Both developed and developing countries have now recognized this rapid growth and are creating a network of institutions and centres in order to support research and development focused towards this industry. ‘The International Forum cum Conference on Information Technology and Communication at the Dawn of the New Millennium, proposed as the Year 2000 activity of the academy and sponsored by Siemens AG, is expected to be a world-class event that will demonstrate important issues of the current IT industry. The wide-ranging applications of IT in the public sectors would include issues such as health care, education, tourism, legal and tax issues as well as customs and immigration, among others. On the other hand, the private sector would concentrate on IT's impact on the business world such as application of E-commerce, multimedia technologies for video and voice transmission (teleconferencing), and Internet services. Since Information technology relies heavily on the communication infrastructure of each country, the Forum would also include advancements in the communication facilities such as wireless communication, satellite services and microwave facilities, and high speed wired infrastructures with optical fibre links, We only had six months to organize this important event. First, the Technical Committee was constituted and then, the Honorary Advisory Committee, the Steering Committee and the Organizing Committee were formed. Full support was received from many Engineering Academies, Universities and Government and Private Sector Agencies. ‘The Conference is jointly organized by the Asian Institute of Technology (AIT) and the National Science and Technology Development Agency (NSTDA) of Thailand. The continuous and full support received from President Jean-Louis Armand of AIT and Dr. Pairash Thajchayapong of NSTDA are gratefully acknowledged. Mr. Stehle Wolfgang and Mr. Hans Eckard Scholz of Siemens AG are thanked for their active support and their guidance in organizing the conference. The conference is held in collaboration with numerous universities in Thailand: Kasetsart University, Saint John's University, King Mongkut’s Institute of Technology, North Bangkok Campus, Bangkok University, Khon Kaen University, Burapha University, Chiang Mai University and Thammasat University. Serving as Chairman of the Honorary Advisory Committee is Prof. Jean-Louis Armand, President of AIT. ‘The Chairman of the Organizing Committee is Dr. Kanchit Malaivongse and the Steering Committee Chairman is Prof. Mario ‘Tabucanon, AIT Provost. The members of all the committees are thanked for their most valued contributions. Within AIT, Prof. Huynh Ngoc Phien, Dean of the School of Advanced ‘Technologies and Dr. Kazi Mohiuddin Ahmed, the Forum Chair of SAT and all other AIT faculty members are thanked for their generous advice. The International Academy is currently housed in the School of Civil Engineering at AIT and Prof, Worsak Kanok-Nukulchai, Dean of the School of Civil Engineering, is thanked for the facilities and services given within the School in undertaking this endeavor. Most of the IT-Giants are participating in this event in one form or another including co sponsoring the activity. These include Oracle, Hewlett Packard, Computer Associates, IBM, Acer Computers, Microsoft, NEC, Nokia, Ericsson, Alcatel, Motorola, Seagate Technology and Sumitomo Industries Ltd., among others. Prof. Cham Tao Soon, the President of the Nanyang Technological University (Singapore) will give the prestigious Bender Lecture. Almost all the best Universities in many parts of the world are in one form or another actively participating in the Conference. From the region, the Nanyang Technological University in Singapore, the Hong Kong Chinese University, the Korea Advanced Institute of Science and Technology (KAIST), theUniversity of Hokkaido in Japan and the Monash University in Australia, among others, are actively collaborating in this important academic and professional endeavor. A well-balanced set of Keynote Lectures will be given by outstanding Academicians and Industry Leaders. The Lectures are of interest to University Teachers and Students, Government State and Private Sector Agencies using IT’ Techniques. Nearly 100 contributed papers are tabled for presentation cover themes such as B-commerce, E-Business, Education, Business Information Systems, Software Tools and Technology, Internet and Mobile Communications. A large exhibition is planned. Excellent Tutorials and Workshops are also included. Each and every activity of the four-day programme would be of immense value to all those who are interested in IT and its applications regardless of the organization they belong to. Written contributions received are contained in three volumes. Volume 1 contains the Bender Lecture and invited papers. Volume 2 compiles contributed papers. Documents received late are included in Volume 3. ‘The Conference will be followed by a Summer School at AIT, also sponsored by Siemens AG and this Summer School is arranged under the able leadership of Prof. R. Sadananda. It is the most valued contribution and assistance received from a large number of persons from all walks of life ‘who conceived and concretized this conference. It is impossible to mention the names of all those who have contributed their efforts. ‘The administrative and organization part of the conference were carried out under the able leadership and dedicated work of Mr. M.LM. Wahab. The following personnel are thanked for their dedicated works. Mr. Thein Htike Oo, Mr. Nguyen Hong Minh, Mr, Jeyendran Mathias, Miss Li Rongmei, Miss Chanidaporn Chaimongkol, Miss Boonjira Intaratood, Miss Yanee Somkaew, Mr. Pratanporn Boonrit, Ms, Han Ying and Ms Anjali Sharma, Khun Kanya Worawichawong from the Development Office was helpful in liasing between AIT and Siemens AG. Ms. Anne Lopez, the Outreach Officer of SCE is thanked for her dedicated. contribution in constantly reviewing and helping in the organization of the conference. Ms. Nualchan Sangthongsathit and Ms. Sontima Pongpanich are thanked for their administrative assistance. It is hoped that similar conferences can now be held in different cities in Asia and elsewhere at periodic intervals. Finally, the volumes of the Proceedings are dedicated to H.E, M.L. Birabhongse Kasemsri, former Chairman of the AIT Board of Trustees, for his most valued and untainted help in the formation of the International Academy and in guiding many of the Activities which are held so far in a successful manner, Kanchit Malaivongse Huynh Ngoc Phien Chongrak Polprasert Kanchana Kanchanasut, John C.S. Tang A.B. Sharma KM. Ahmed Phan Minh Dung Qi Yulu Vilas Wuwongse Worsak Kanok-Nukulchai AS. BalasubramaniamTABLE OF CONTENTS 1. The Impact of IT on University Education............ Cham Tao Soon 2, Information Systems and Web Information Systems: A Methodological Perspective. Colette Rolland 3. VLSI Design Education for Undergraduate and Graduate Students. Yoshikazu Miyanaga 4. Designing 3G Applications for Smartphones and Wireless Data Terminals... Gareth Loudon 5. E-Business in South East Asia..... 4B Mark Gabbot 6. Application Service Providers (The Latest Wave in E-Commerce) v...-sseess 53 Robert Kheng Leng Gay 7. 30 Years of Software Engineering: From Art to Industry?..... ‘Anne-Marie Hugues 8. Informatics: A Truly Interdisciplinary Science - Prospects for an Emerging World -.. Dines Bjomner 9. Algebraic Switching Theory and Broadband Applications... 85 Part I: Layout Complexity of Multi-Stage Interconnection Networks Shuo:Yen Robert Li & Hui Li 10. XML Declarative Description ~ A Unified Information Representation. 9 Vilas Wuwongse 11. Knowledge Based Economies: Some Challenges for 2000 and Beyond. 105 Jack Wood 12, Heterostructure Microelectronics in Information Systems. eold5 Saburo Takamiya & Kouichi liyama 13. Wavelength Division Multiplexing (WDM) in Optical Communication Systems...........131 Pak L Chu 14, Building an Information Society for Us All... 1B Nick MoorePresentation by Dr Cham Tao Soon President, Nanyang Technological University = “The Impact of IT on University Education” The Bender Lecture at the Asian Institute of Technology -1 August 2000 = Nae __ _ Introduction The speed at which new knowledge is acquired is astonishing. Doubling of human knowledge has shortened from about 100 years to a matter of every few years. In addition, this new knowledge is made easily available to everyone due to the advancement in info-communication. The powerful search engines of the Internet have given us immediate access to huge amount of data at the click of the mouse. In this environment, the basis of the economy has shifted from the production of goods to the production of knowledge. Leaming takes on a new role. In this modern world, ‘The illiterates of the twenty-first century are not those who cannot read and write, but those who cannot learn, unlearn and relearn” — an analysis by Alvin Toffler. The Learning Revolution These changes have deep implications for us as learners. Firstly, the speed at which new knowledge is created and how it is delivered place us in a new dimension in learning. There is no way anyone can hope that what he has learnt will stay relevant for long. The need to continue learning so that we can keep pace with the new knowledge has become critical. Secondly, in the new economy, it is not only the ability to access and process information that is important, but also the ability to create new knowledge. Therefore, creativity and innovation have to be emphasised. Thirdly, learning using technology is a new experience. Learning the technology is different from maximising the advantage of having access to so much information. It is about the ability to navigate through the mass of information and discriminate between what is relevant and what is not ASEAN had been active partners with the MNC's in the global economy. We have contributed physical resources and human capital in manufacturing, services and commodities. Our competitive advantage had been based on responsiveness, quality and low cost. This is now threatened by the transition to a global Knowledge Based Economy. Agility is more importantthan size, history or track record in the new economy. The info-communication revolution based on the Internet is both an enabler as well as a leveller. The pressure to respond has resulted in demands for new competence, skill upgrades, training and productivity. © Current employees must upgrade continuously to be relevant. Experience in some function may be increasingly replaced by new knowledge requirements. There are a number of implications for the university due to these rapid changes. First, the half-life of knowledge continues to diminish rapidly. An undergraduate education can no longer be seen as equipping graduates for a career that will span three or four decades. Our graduates will require constant training and retraining throughout their working lives. This training will take on a number of different forms, including retuming to the university for further education. Secondly, the sources of knowledge are much less well defined than in the past. The Internet and the World Wide Web now provide information content in a convenient and low-cost fashion that is without parallel Shakespeare's Hamlet is modernised with a new quote “To e or not to e, that is the question. There is ‘e’ in everything, including e-learning. As a result, the role of the professor as provider of knowledge diminishes day by day just as the Tole of the professor as guide and mentor, becomes more and more important. Thirdly, the graduate will increasingly work as a member of a number of disparate teams. Indeed, the current knowledge worker may be simultaneously a member of several teams, within the enterprise, across enterprises, and in their profession and community. Many of these teams may be virtual teams where knowledge workers co-operate across national boundaries and time Zones using email, voice and video-conferencing. Fourthly, in a global environment, the role of national cultures is more and more important. Not only must the university be seen as a pivotal instrument of nation building and vital store of our own national cultures, it must also communicate to our students and graduates the consequence of other cultures with which they will interact throughout their working lives. Now, of course, information technology cannot be a solution to all of the many challenges that these changes bring to the university. For example, information technology can provide some introduction to students of other cultures, but not at the level as that provided by having foreign students sitting in their classes or by travelling to other countries for a semester or two, as we are doing in our own institution. Information technology can, however, go some way to assisting the university meet the challenges of the new knowledge-based economy. Investment in information technology should not, however, be treated as a fad or as an outcome of some marketing or “brand building” exercise. | believe that it is important to recognise that investment in information technology can only be justified if there are educational payoffs. | will now address some of those possible payoffs.THE FUTURE UNIVERSITY Earlier, | noted that the half-life of stores of knowledge continues to diminish. This applies to engineering, medicine, science and business disciplines. Our graduates are increasingly required to prepare themselves for several complete changes in their work patterns. An engineering graduate may, for example, spend only four or five years after graduation performing the type of technical tasks that their university studies prepared them for, before moving into sales or marketing and then on to general management and perhaps back into managing technical team. At each stage of their development new general and technical skills are required. In the past, a graduate might have worked for a single enterprise and the enterprise may have mapped out an educational program. Now, however, the graduate is more likely to sequentially work for a number of different firms or to set up their own enterprises. The responsibility and commitment that the employee feels to their particular employer is consequently reduced. Some commentators have described this more flexible environment as the “free agent” effect. In such a world, the graduate takes on increasing responsibilty for the development of his or her own learning. The task for the University is, then, to imbue their undergraduate programs with the spirit of continuous learning while at the same time ensuring that their students receive a rounded professional education. Information technology can assist with this process. Firstly, IT can facilitate the provision of content that can be delivered at any time and at any place. Instructional delivery needs no longer be limited by the constraints of the scheduled lecture, laboratory and tutorial. Secondly, content can employ multimedia-rich resources that can be delivered by CD-ROM or network. Thirdly, IT can provide immediate feedback and interaction in a manner that is difficult to achieve with traditional modes of instruction. Fourthly, the provision of content need not be limited to the institution's own faculty. Indeed, as | will discuss shortly, the best scholars from different institutions may deliver content to students around the world. For example, our University is offering an MSc in Innovation and Manufacturing System conducted jointly with MIT. Some of the courses are shared with students located in MIT through tele-conferencing and Internet 2 The spirit of enquiry and the development of the students’ own construction of knowledge can be facilitated by the more general resources provided over the network. My own institution has made considerable investments in recent years in developing electronic knowledge resources. Our Gateway to Electronic Media Services (GEMS) provides to the student whether he or she is in the library, hall of residence or at home a range of databases and information resources that would have been unthinkable only a few years ago. Indeed, my colleagues tell me that it is difficult for them to not only keep up with the range of resources that our library now provides through GEMS but to plan how those resources are to be employed in the educational process. IT can also assist in the way in which students interact with each other and with their professors. Asynchronous learning networks, includingemail and discussion forums, can bring students closer to each other and to their professors. There is a widespread view that IT can isolate students from each other and from the wider society. When one sees youngsters playing Quake Ill for hours after hours in cyber cafés it can be seen that this perspective has some foundation. This view of IT as a tool of isolation need not, and in the context of a university education, should not, be upheld. University educators should employ IT as a means to facilitate small and large group interaction. In my institution we are increasingly emphasizing the role of the group in problem solving activities. In our undergraduate business and communications studies programs, for example, students are constantly involved in group work. It is ‘common for students to manage their ongoing participation in four or five groups. IT can facilitate this complex management task by providing tools for research, group communication and writing activities and for the submission and dissemination of group results. Communication between students and professors can also be facilitated by IT. There are, of course, considerable power differences between professors and students. While we seek active participation of students in tutorials and laboratories, there will always be barriers that arise from differences in status, age, knowledge and experience. The relative anonymity that IT brings can serve to break down some of these barriers. Further, IT can facilitate ‘communication to large numbers of students. If one student asks a professor a question, it is highly likely that many other students have the same question. Or perhaps, more likely, the other students realise that they should have asked that question when they see the question and the answer. IT can provide a mechanism by which an individual question or comment from a student can illuminate difficult issues for the complete student body. At NTU, we are in the relatively early stages of using IT for these Purposes but | am hopeful that we can use IT to make interaction and communication more productive. | can say, however, from my observations gained that the information technologies that are in most common use by students are not the PC or laptop or the Internet. The technologies most in evidence, in contrast, are the cell phone and the pager. Students make intensive use of these technologies to plan their group activities; to discuss their tutorial homework and their extracurricular activities. 1 expect with the introduction of the 3G cell-phones using the WAP, these activities will further intensify. There is a lesson here for us as educators — students seek out human contact and will use whatever technology is most appropriate to further such communication. Indeed, one of the commonly held beliefs about the application of IT to learning is that it comes at the cost of human interaction between professor and student. There is a tendency in some quarters to view IT as a cost-effective substitute for the professor and that all students will do during their university career is to interact with a computer. As an educator, | would be concerned if this were to be the case. Information technology will, | believe, change the mode of interaction betweenprofessor and student. The professors as a receptacle of knowledge that is transferred to the student in the lecture context will give way to different forms of interaction, The venerable lecture is unlikely to disappear from the university setting. — However, rather than being the primary source of content and knowledge for the student, the lecture is increasingly likely to provide context and structure to the knowledge that is more productively made available on CD- ROM and on the network. The nature of small group interaction between professor and student is also likely to change. Increasingly a professor can assume that his or her students have access to a body of knowledge in electronic form, complemented by assessment and reinforcement tools. The role of the professor in tutorials and laboratories moves more and more to that of a guide and mentor and someone who challenges the student to place their learning in context. 1 am not suggesting that public institutions such as my own will become like Cambridge or Oxford with its tutorial system overnight. | am, however, suggesting that Information Technology will free up the professors to engage with their students on a different level than is now the case. PARTICULAR ISSUES FOR ASIAN INSTITUTIONS | turn now to some of the implications of information technology for public universities in the Asian region. There are two major issues. First, the application of information technology to teaching and learning gives rise to many questions for the design of educational processes and to the management of the university budget. Second, just as we have seen the implications for global business of electronic commerce, so we are seeing the arriving of global universities that are marketing their brand and intellectual property into our region. | will discuss each of these challenges in turn. Managing the design and implementation of educational programs is a complex task. Unfortunately, there is no standard handbook for applying information technology to teaching and learning processes. Change may be difficult for professors who grew up in a period before there were computers and who may be uncomfortable with the changed roles that they may be required to take on as developers and users of information technology and in the way that they interact with their students. In contrast, many of our undergraduates have been using a variety of information technologies throughout their lifetime; a generation that Don Tapscott categorises as the N- or Net-generation. They are familiar with gaming, email and the Web in a way that most professors are not. Universities everywhere are grappling with how best to integrate information technology into the curriculum and how to provide faculty with a new literacy. University management is also grappling with the budgetary implications of information technology. It seems that every visit of my Director for Information Technology Services brings another budget request for staff, hardware, software, local area networks or Internet connectivity. NTU has recently embarked upon the development of a campus-wide wireless Local Area Network covering 200 hectares that will enable faculty and students to connectto the network whether they are in the tutorial room or the canteens. This represents a significant investment on the part of the University and one where it is difficult to point to immediate payoffs. The second challenge is the increasingly global competition in the provision of higher education by well-resourced institutions and consortia from countries outside our region. Just as IT and ubiquitous and low-cost communication networks have brought major changes for commerce, so are they bringing change to higher education. It is not that we in this region have been immune to competitive pressures in the past. Singapore is, for example, the host to more than 40 offshore MBA programs from Australasia, Europe and North America. The nature of the competition caused by these offshore Programs is changing and changing rapidly. In the past it may have been appropriate to question the quality of such programs, or in some cases the standing of the institutions that present the programs. Now, however, we are seeing some of the most prestigious universities taking on |T-facilitated distance education. One of the longer established global business programs is the Global Executive MBA, or GEMBA, program offered by the Fuqua School of Business at Duke University in the USA. The GEMBA program uses a mixture of techniques to achieve its educational objectives. The GEMBA program emphasises the importance of face-to-face interaction and intensive periods of residential study at different locations around the world. Not only does this build skills in areas of content but also develops the intimate personal contact and trust between faculty and students and between students and students that is so important, all due to the IT-facilitated communication that takes place in the times between the residential study periods. Students are provided with laptops and access to a variety of standalone and networked knowledge and communication tools. They interact in both synchronous and asynchronous modes, made possible by IT. The GEMBA program is known as being one of the more expensive executive MBAs in the marketplace. Its design that enables busy executives to continue to study wherever they are in the world; global approach and mix of face-to-face and network enhancement has meant that it has been able to successfully compete against other executive programs. The danger is that programs of this nature may give rise to a homogenised view of the world that comes from North American or European sensitivities and does not reflect the realities or particular needs of our region. We are not able, of course, to prevent competition coming from overseas institutions. We can only overcome such competition by producing a higher quality educational product; by utilising information technology in a way that our competitors do; by making the programs directly relevant to an Asian audience and by ourselves entering into consortia with regional and global players.PROBLEMS AND CHALLENGES Information technology in pursuit of learning is not a panacea. More than twenty years of educational research into the application of information technologies to teaching and leaming does not provide clear evidence that IT brings measurable benefits to learning or the acquisition of skills. Much of the research that has been conducted has been contaminated by the involvement and enthusiasm of the researcher, commonly known as the Hawthorn effect. It is difficult to evaluate the effect of information technology on the complex interrelationships that go to make up learning | think it is necessary to sound a waming note on the application of IT to the learning process. | spoke earlier of the increasing need for our students to be prepared for a life of continuous learning and for them to construct their own knowledge and knowledge frameworks. This does not mean, however, that we are prepared to accept knowledge at the level of a television situation comedy. Acquisition of learning at the university level requires integration of different techniques and bodies of knowledge. This requires intensive application to the knowledge construction task. One of the concerns with applying information technology to leaming is that we may give the impression to the student that a task is easier than is actually the case. The use of information technology in learning may also give rise to a level of superficiality. The hypertext format of the World Wide Web in particular facilitates rapid movement from page to page and from subject to subject. This gives rise to superficial level of knowledge as students can scan pages and move quickly to the next subject. This is the equivalent in learning of channel surfing between television channels. It behoves us as educators to ensure that students are immersed in learning rather than being at a surface level. | cannot stress enough the importance of supplementing intensive use of information technology with equally intensive face-to-face communication between professor and student. The old issue of user authentication that challenges e-commerce applications rears its head in e-learning. Is that the student's work or that of another's? Such issues of plagiarism have yet to find definite solutions. Is the student taking the remote test that he is supposed to be? Could it be his colleague, sibling or paid expert? Biometrics provides a partial answer — it might be him and someone else doing the test together. Perhaps the traditional examination hall has to remain. One innovative way to overcome this is by clever assignment design. In this case, challenging projects are given to students to evaluate their independent thinking beyond their (as well as the web's) knowledgebase. Related to this is the copyright and intellectual property (IP) concerns. The jury for this tricky legal issue is still out in their deliberations. At the end of the day, while technology brings evolution to the educational processes, the academic community must remember the cardinal principle ; that educational change is not, and should not be, technologicallydriven — but itis, and always has been technologically enabled. For this reason, we coin the word feachnology, technology with an “a". They should , in a single word, balance the progress, goal and vision of the central objectives of the teaching function enabled by technology. The tools have changed. Staff and, to a lesser extent, students have to be trained to use them. Finally, while IT is a useful tool to support knowledge acquisition and interaction, there is | believe still no substitute for music, poetry, theatre and reading as well as involvement in sport and other physical activities and intensive informal discussions on the issues of the day. We will have failed as. educators if all that our students do is interact with a computer. A piece of interesting news was Michael J Saylor, President and CEO of MicroStrategy Inc made the educational headlines early this year when he pledged to donate US$100M (prior to his US$6B single day loss) to establish an online university with the motto of “free education for everyone on earth, forever’. Interestingly, a reply in the pages of Washington Post by Michele Toleda Myers, a President of a US College, provided a good reminder to that vision. Myers said, “If education were only as simple as reading, then libraries would have replaced schools long ago. We educators are in the business of forming minds — not just fling them’. ‘es:50:28.08-2000INFORMATION SYSTEMS AND WEB INFORMATION SYSTEMS : A METHODOLOGICAL PERSPECTIVE, Colette Rolland Université Paris] Panthéon Sorbonne CRI, 90 rue de Tolbiac 75013 Paris, France rolland@univ-paris| fr ABSTRACT The Development of Information Systems (ISD) is a complex activity that requires methodological support. Current ISD methods are predominantly following an object driven paradigm that allows us to capture the static as well as the dynamic dimensions of real world objects into information objects. However, Web Information Systems have facets such as information presentation, user profile, navigation structure etc. which pose new challenges to ISD methods. The keynote talk will provide an overview of current ISD methods and discuss the new challenges raised by the development of Web Information Systems. This paper is a brief overview of the key points developed in the talk. INFORMATION SYSTEM DEVELOPMENT In this section we briefly introduce the two-phase organisation of the process for developing information systems namely, Requirements Engineering as a prerequisite for System Engineering. We then, use the [Olle88] framework for classifying IS conceptual models that traditionally support the requirements engineering phase and comment their characteristics. We discuss the limitations of conceptual models as supports for ensuring the match of system functions to organisational requirements and then, introduce more recent requirements engineering models. Finally we comment the evolution of process models to support the modelling process. ISD organisation Traditionally Information System Development (ISD) has made the assumption that an information system captures some excerpt of world history and hence has concentrated on modelling information about the Universe of Discourse [Olle88). Thus the IS development can be organised in two phases as shown in Figure 1. The first phase where conceptual modelling is carried out aims at abstracting the specification of the required information system i.e. the conceptual schema, from an analysis of the relevant aspects of the Universe of Discourse about which the user community needs information [Dubois89]. The succeeding phase, that of system engineering, uses the conceptual schema to design and implement a working system which is verified against the conceptual schema.7 pawerae oF Domain knowtedge sequistion and modetlng a ee CaS Figure 1. Two-phase organisation of system life-cycle Vatidation CONCEPTUAL MODELLING Classification framework of conceptual models The information systems community has developed a large number of conceptual models for representing conceptual schemata. This variety has arisen because of the need to capture as many aspects of real world semantics as possible. Given this plethora of models, it has been found necessary to develop frameworks for classifying and understanding these. One framework which classifies models based on the perspective adopted to view the Universe of Discourse was developed by [Olle88]. It organises models into the classes of process- oriented, data-oriented, and behaviour-oriented models. In Figure 2, this framework has been shown as defining a three-dimensional space within which conceptual models can be positioned. oats in Syatemteepprosches Framework of understanding & “state oriented” (Cia Aiseeicto uy MERE aves Obincpatforny gsA0M CM fREMORA apphoaches ERAE snot Sytem ppronctes = “change oriented” | 8usness Clas Ie Process: cartesian fenrown ataties i Saat oy, PETRI Nets Figure 2 : The three dimensional framework for classifying conceptual models The three dimensional framework highlights the fact that information systems can be looked upon in three different ways. When seen as process-oriented, an information system is a function in an organisation which returns some information. When seen as data-oriented, information systems are viewed as mirroring the information contents of organisations and it is expected that the information system would be a supplier of this information, Finally, in the behavioural perspective, an information system is an artefact which handles interesting events that occur in the organisation by performing one or more functions. These functions modify the information contents of the information system which are again available for manipulation through events. These different views naturally lend themselves to specific kinds of treatment. Thus, when the information system is viewed as a function in the Universe of Discourse, then during analysis, the components of this function are discovered. This is because the function may be very complex and needs to be broken down into its functional elements to understand it better. If any of the functional components are themselves complex then, they are 10decomposed recursively till simple, well understood functions are reached. Clearly, this results in a hierarchy of functions rooted in the original function. Whereas this hierarchy identifies the functional components of the information system function it does not establish an inter-relationship between these components, i.e., which function receives data from which function and sends data to which one is not articulated. This is done by using conceptual models for building data flow diagrams. It can be seen that the process-oriented perspective views information systems as processors of information. In contrast, the data oriented approach looks at an information system as mirroring the information contents of the real world, as a storehouse of information. Since information is to be kept about real life things, an identification of all these relevant ‘things’ coupled with their abstraction as information carrying entities is carried out. The abstracted entities and their inter-relationships are then represented as a conceptual schema As the mirrored world changes, so the information system must reflect these changes, Therefore the information system is seen as a data manager, maintaining and delivering information at all times. Finally, in the behavioural perspective, the attempt is to identify the interesting events that occur in the real world, the information affected by their occurrence and the functions that cause this effect to be felt. For this, three things are done (a) Real events are abstracted into information bearing events, (b) Real world things are abstracted (as in the data Perspective) into information bearing entities and relationships, and (c) Functions to be invoked to carry out the effect of the event are identified and associated with it. It can be seen that the behavioural view promotes a transaction management view of an information system. Over the years, the usefulness of having three completely different perspectives with little integration in them has come to be questioned. Two distinct trends towards integration emerged. The first was the development of object-oriented conceptual models, the majority of which integrated together the process and data-oriented perspectives, though some conceptual models that also integrated the behavioural one were developed [Brunet90], [Desfray94], (Martin92]. The second was a trend towards ‘loosely connected’ conceptual models which consisted of a set of conceptual models, cach according to a different perspective. Therefore, the Universe of Discourse was conceptualised as individual but connected conceptual schemata. This inter-connection was seen in the Yourdon approach [Yourdon89] in the mid- eighties which loosely connected the data flow, ER modelling and state transition diagram techniques. It was also seen later in OMT [Rumbaugh91] which integrated an object-oriented model with data flow diagrams and event modelling. This movement led to the unified Modelling language, UML as a denotational standard for expressing conceptual schemata as well as design and implementation schemata, ments Engineering models In the traditional view of engineering information systems through conceptual modelling, the focus is on producing a formal specification of the system to be developed. This specification concentrate on what the system should do, that is on its functionality. Such a specification acts as a prescription for system construction. It seems that this leads to a number of failures in system development. Indeed, a number of studies show [Lubars93; McGraw97; Standish95] that systems fail due to an inadequate or insufficient understanding, of the requirements they seek to address. Further, the amount of effort needed to fix these systems has been found to be very high [Johnson95]. To correct this situation, it is necessary to address the issue of requirements elicitation, expression and validation in a more focussed ‘manner. The expectation is that as a result of this, more acceptable systems will be developed in the future. Req 11In tackling the above problem, the area of requirements engineering tries to go beyond the functionality based view of conceptual modelling. Requirements engineering extends the ‘what is done by the system’ approach with the 'why is the system like this' view. This why question is answered in terms of organisational objectives and their impact on information systems supporting the organisation. In other words, information systems are seen as fulfilling a certain purpose in an organisation and requirements engineering helps in the conceptualisation of these purposeful systems. This has two implications (a) elicitation and validation of the requirements of a system is done with respect to their purpose in organisations and (b) only organisationally purposeful systems are conceptualised. To deliver the foregoing, requirements engineering models have being developed that can be classified into : - goal-driven approaches and, - _ scenario-based approaches. Goal driven approaches model organisational objectives so as to relate them to the functions of the system. In this sense, they aim at the conceptualisation of purposeful systems only. They contribute to the interpretation of requirements before they are understood and before they are transformed into system function specifications. Thus they support conceptualising purposeful systems. The broader view of a requirements specification that they advocate is going beyond the classical conceptual schema describing the system functionality. It includes enterprise modelling which represents the Why part of system requirements. Enterprise modelling as developed in the F3 project [Bubenko94] is an example of such a goal-driven approach to requirements engineering. Enterprise modelling was further refined in the EKD method to support change management [Loucopoulos98; Rolland97; Kardasis98; Rolland98}. In the KAOS approach [Dardenne93], the emphasis is on supporting formal refinement of high level goals into system constraints. Although generic models are advocated, goal modelling and refinement have supplied simple guidance via heuristics[van Lamsweerde95]. The I* approach [Yu94a, b, c] creates models of the environment of the system that emphasise agents and their relationships. Their strategic dependency and rationale models allow tracing of dependencies between agents, goals and tasks and support reasoning to identify trade-offs between functional requirements and non functional requirements [Mylopoulos92} Independently of goal modelling, an alternative approach to RE, the scenario-based approach [Jacobson95] has been developed. Scenario based approaches, by focussing on the users’ view points, help in modelling purposeful system usage from which useful system functions can be derived. Scenarios provide dynamic meaning to goals whereas goals provide the intentional setting within which scenarios find meaning. By capturing examples, scenes, narrative descriptions of contexts, use cases and illustrations of agent behaviours, scenarios have proved useful in requirements elicitation in a number of ways : to elicit requirements in envisioned situations [Potts97], to help in the discovery of exceptional cases, to derive conceptual object-oriented models, to understand needs through scenario prototyping and animation, to reason about design decisions, to create context for design [Kyng95] and so on. The underlying reason for the popularity of scenario-based approaches seems to be that people react to descriptions of real happenings and real things. This reaction helps in clarifying requirements expected of systems. Thus, the scenario school argues, that typical scenarios are easier to get in the first place than goals. Goals can be made explicit only after deeper understanding of the system has been gained. Scenarios have been developed for different purposes with different contents, expressed in different levels of abstraction and with different notations. In [Rolland98b] the 12reader will find a framework to classifying and comparing scenario-based approaches and an analysis of 12 significant scenario-based approaches. It might also be noticed that in order to overcome some of the deficiencies and limitations of goal-driven and scenario-based approaches used in isolation, some proposals have been made recently to couple goals and scenarios together. Goals have been considered as contextual properties of use cases and as a means to structure use cases. The goal scenario combination has been used to operationalise goals, to check whether or not the current system usage captured through multimedia scenarios fulfils its expected goals, to infer goals specifications from operational scenarios and to discover new goals through scenario analysis. ‘As an example of an approach which combines goal modelling and scenario authoring consider the CREWS-L'Ecritoire approach [Rolland97; Rolland98a] developed within the CREWS ESPRIT project. CREWS-L'Ecritoire uses a bi-directional coupling allowing movement from goals to scenarios and vice versa. The complete solution is in two parts : when a goal is discovered, a scenario can be authored for it and once a scenario has been authored, it is analysed to yield goals. By exploiting the goal-scenario relationship in the reverse direction, i.e. from scenario to goals, the approach proactively guides the requirements elicitation process. In this process, goal discovery and scenario authoring are complementary steps and goals are incrementally discovered by repeating the goal-discovery, scenario- authoring cycle. ISD process models This brief overview of ISD methods would not be completed without considering the second dimension of any method : the process dimension. This is dealt with in the following, The conceptual modelling community emphasised the product aspects of systems at the expense of the process employed to deliver the product. Thus, the structure of the conceptual schema, its completeness, and consistency etc. was more important than how it was developed. Early process models were activity based. They looked upon the process as consisting of a set of activities which could be decomposed into simpler ones and which were linearly ordered. Every successive activity was to be performed after the completion of the previous one. Such process models are known to be restrictive [Wynekoop93] because they assume (@) that it is possible to pre-define the development path that can be taken through the activities of a process model. Thus, they restrict the creativity of the developer in choosing a path specific to a given situation. (b) that each conceptual schema is built afresh and therefore there is no need to keep track of the processes that built them. (©) the ‘upon completion’ rule which prohibits movement to an activity later in the order or backtracking to one earlier in the order. (@) that the relationship between an activity and the product built by it was not interesting, Later, a number of other more flexible process models were built. Yet, by and large, conceptual modelling continued to follow the activity based approach to process models i.e. the Waterfall model [Royce70] In contrast to conceptual modelling that largely ignored the development process, requirements engineering has explicitly considered the issue of the process support to be provided. Two important issues arise : 1, How can attention be channelled to deal with the real productive tasks of requirements engineering? In other words, it is necessary to guide the requirements engineering process to concentrate on discovering goals, scenarios etc. 132. How can one lear from past practice? That is, if some decisions were taken in a given situation in the past then how can one benefit from experience with that? Thus it is necessary to keep a trace of past decisions. These two aspects of the requirements engineering process, namely guidance and tracing led to guidance and trace models embedded in computer assisted tools to support the RE process in a semi-automated manner. Guidance Some experience in guidance exists in software engineering where guidance was classified as active or passive (Dowson94]. The former was focussed on ensuring that the development process employed was an instance of the process model and consequently, guidance was directed towards process model enforcement. The latter was concemed with an identification of what could be done next in the development process. In [Feiler93] passive guidance has been defined as the generation and subsequent presentation of the set of legal steps that were available at any moment in the development process. One out of these could then be selected as the task to be done next, The software engineering view is that active guidance should be provided. Thus, guidance cannot be provided without an adequate process model. Existing process models do not seem adequate to requirements engineering as they prescribe a predefined plan of actions. Activity-oriented process models (Royce70] come from an analogy with problem-solving and provide a frame for manual management of projects. This linear view is inadequate for methods which support backtracking, reuse of previous designs, and parallel engineering Product-oriented process models [Humphrey89; Finkelstein90;, Franckson91] represent the development process through the evolution of the product. They permit design tracing in terms of the transformations performed and the resulting products. Finally, decision-oriented models integrate more deeply the semantics attached to evolutionary aspects. The notion of design decision facilitates understanding of the designer's intention, and better reuse of results [Potts89] but the flexibility and situatedness of requirements engineering processes is not adequately handled in existing decision-oriented models. The importance of situatedness in process modelling is also acknowledged by the software engineering community where it was found that departures from the process model occurred in actual practice. A concerted effort was put in to allow process models to respond to these departures. One approach was to assume prescriptive models and then, modify them, to accommodate real processes. This modification could be achieved in two ways. First the extent of deviations from the prescription that could be allowed was modelled as constraints. Any actual deviation that satisfied the constraints was therefore manageable and the process enactment mechanism could handle it. This way of handling deviations took the prescriptive approach to its logical conclusion: it prescribed the deviations allowed in a prescription. The second way of handling deviations was to allow changes to be made in the prescription as and when they are needed. Thus, a level of dynamicity is superimposed on the basic prescription. In contrast to this, the requirements engineering community recognised that the core of their task was the generation and exploration of alternatives from which the right one is selected for the situation at hand. This can be seen in the IBIS process model [Potts89] where a number of alternatives for resolving an issue were generated. This process model is at a very high level of abstraction and had to be buttoned down to real methods and tools. The contextual model [Rolland91; Rolland95; Poh!96] attempted to do this. A context was defined here as the application of an intention to a given requirements engineering situation. It organised requirements engineering methods as a set of contexts of three kinds, executable, plan, and choice contexts respectively. A Choice context groups together all possible alternative ways of meeting its intention. These alternatives were themselves contexts thus 14leading to a hierarchy of alternatives. A plan context is a collection of simpler contexts such that their execution, in the various possible orders prescribed in the plan context meets its intention. Finally, an executable context is one which can be directly executed to meet its intention (and is atomic in this sense). It can be seen that the contextual model attempted to reconcile process prescription with alternative generation, the former through plan contexts and the latter through choice contexts. Another attempt to root the notion of alternatives in methods was made in the decisional approach [Prakash97] Tracing In the requirements engineering community there is no longer the question whether traceability is a useful thing or not. Capturing and maintaining traces is seen as an essential activity to be performed during requirements engineering and standards such as [DoD-2167A; IEE-830] mandate that requirements traceability be practiced. A comprehensive overview of possible usage of trace information and the expected benefits can be found in [Gotel94}, {Ramesh93a] and (PohI96]. These reports indicate that requirement traceability is a vital component in implementing a quality system, essential for consistent change integration, leads to less errors during system development, plays an important role in contract situations, and improves system acceptance. Process traceability can be divided into three parts [Poh196] : * Process execution traceability, i. the recording of data that enables the reassembly of the sequence of steps of a process. * Product evolution traceability, i.e. the recording of data that enables you to see how the product has evolved during the process. * Traceability of the relationships between process execution and product evolution. The pivotal goal of process traceability is to enable tracing of the requirements produced during the RE process. On one hand, traceability from the requirements specification through design to implementation and vice-versa is needed to understand the rationale of the implemented system. On the other hand, the process leading to the requirements specification must be traceable to understand the rationale for the requirements themselves. The former is referred to as post- traceability whereas the latter is called pre- traceability [IEEE-830] Product traceability is available in some methods like Class/Relation, OOSE and rAdar. Post-traceability is supported by some commercial tools like RT from Teledyne Brown Engineering, RMT from Marconi Systems Technology , and RDD100 from Ascent Logic. Pre-traceability has been investigated more recently [Gotel94; Pohl96; Ramesh93b; Ramesh93] An interesting framework for requirements pre-traceability was provided by Pohl [Pohi94] who described the requirements engineering process in a three dimensional space assuming that there are three major facets of the RE process, namely modelling the requirements in a more complete manner, modelling with more formality, and more consensus among stakeholders. The trace of the requirements engineering process is modelled as a path within the three dimensional space starting from an initial incomplete, informal specification representative of individual viewpoints and ending with the desired output which is a complete, fully agreed and formally described specification of the intended system. Capturing the RE process trace and thereby establishing requirements pre-traceability ‘means recording information along each of the three dimensions, on the relationships between the three kinds of information and relating those to actual process performance. 15WEB INFORMATION SYSTEM DEVELOPMENT Because of the rapid development of the Web technology and of the increasing interest of users and developers, the notion of a Web site is moving from a set of HTML pages fo Web-based Information System (WIS). A WIS is an Information System providing facilities to its to access complex data and interactive services through the Web. E-business applications, Intranet systems, CRM and supply chain applications are examples of WIS. Despite this rapid evolution (or because of this), WIS development is essentially “ad hoc’. Developers often consider Web development as a media manipulation and presentation creation rather than traditional IS development. Thus, WIS development does not follow the well established engineering principles and consequently, it is difficult to ensure the quality of the resulting product. Another important factor is that the Web users’ community is very heterogeneous. Many of the so-called ‘Internauts’ navigating on a site might have different goals and different backgrounds and knowledge. This leads to usability problems such as “loss in Hyperspace" and "cognitive overload" which have been reported in the literature. WIS development is more complex than IS development and raises many new issues such as presentation issues, user profiling, dynamic adaptation of the format and informational content presented to a user, navigation support etc. Over all, there is clearly a need for developing WIS development method that can provide a disciplined way-of-working to ensure quality WIS development. There is some concerns in the literature about the problems that can occur if WIS development remains ‘ad-hoc’ [Zelnick98], [Gibbs94] As an example of such effort to define a WIS development method, let us briefly introduce the AWIS-Method (AWIS-M) [Gnaho99]. AWIS-M support the analysis and design of Adaptive Web-based Information Systems (AWIS). The key aim of the AWIS method is to adapt the inforrtation provided and the services offered by a WIS to the individual needs and preferences of its users. As shown in Figure 3, the AWIS-M considers the creation of an Adaptive Web-based Information System according to four perspectives : - the management of the informational contents, - the definition of the navigational structure, - the definition of the user interface and, - the identification and description of potential users and of their goals. Figure 3. The four perspectives of Web-based systems development Each of these views is supported by appropriate models which allows the WIS engineer to design the system considering each view in isolation from the others. 16- Models that support the Users & users’ Goals view capture (a) information about the potential user such as their background, knowledge, preferences etc. in order to define user categories and (b) usage goals of these potential users. = Models related to the Jnformation view are used to model the WIS information contents, i.e. the domain knowledge that the WIS will store and retum to its users. - Models that support the Navigation view deals with the presentation structure of the WIS informational contents. They help structuring the hyperspace in a net of nodes and links among them, - Finally, models associated to the Interface view deal with the formatting of pages associated to the navigation structure. Obviously, traditional IS development is mainly concemed with the second view and to acertain extend, with the fourth one. Besides, whereas the three last views are mentioned in a number of WIS development research projects [Femandez98], the first one is an original contribution of the AWIS-Method. By capturing the user profiles, the method is able to define user categories and to tune the presentation of the WIS contents according to the specificity of the user profile. Besides, capturing the user goals, the method is able to define guidelines for navigating in the hyperspace to optimise the satisfaction of the user goal It shall be noticed that identifying the four perspectives help mastering the complexity of a WIS development. There is some independence between the view that allows the WIS engineer to model a view in isolation from the others. In addition, the WIS engineer can define several navigational structures for a single informational contents, or present the same navigational structure in different ways. Therefore, the goal is not to create a single Web Information System, but a set of views of the same system, each of them being aligned to the individual needs and interests of a WIS user. CONCLUSION The paper is an overview of the keynote talk that will be delivered during the Conference. It focuses on a methodological perspective to both IS development and Web IS development. The key part of this overview is a state of the art on ISD methods, considering both their product aspect and their process aspects. The former deals with conceptual and requirements engineering models whereas the latter is concerned with process models to guide and trace the system development process. The paper introduces the new challenges posed by Web ISD and discusses some of the key issues. REFERENCES Brunet J., Cauvet C., Lasoudris (1990), Why using Events in a High-level Specification, in Proceedings of the International Conference on Entity-Relationship Modelling, ER90, Lausanne, Switzerland. Bubenko J., Rolland C., Loucopoulos P., De Antonnellis V. (1994) Facilitating Fuzzy to Formal Requirements Modelling, Proc. Int. Conf. on Requirements Engineering (ICRE), Colorado Springs, US. A, Av. Laamsweerde and Fickas S. (1993) Goal Directed Requirements mn, Science of Computer Programming, 20 (1-2), pp3-50. Desfray, P. (1994) Object Engineering, the Fourth Dimension, Addison-Wesley/Masson,, DoD-2667A Military Standard (1988) Defence System Software Development, Department of Defence. Dowson M., Femstrom C. (1994) Towards requirements for Enactment Mechanisms, Proc. European Workshop on Software Process Technology. 17Dubois E., Hagelstein J., Rifaut A. (1989) Formal Requirements Engineering with ERAE, Philips Journal of Research, Vol 43, No 4 Feiler PH. and Humphrey W.S. (1993) Software Process Development and Enactment: Concepts and Definitions, Proceedings of the. Second Intl. Conf. On Software Process. Femandez MF., Florescu D., Levy A.Y., and Suciu D. (1998) Catching the boat with Strudel: Experiences with a web site management system, in Proc. Sigmod’98, pages 414-425, Seattle. Finkelstein A., Kramer J., Goedicke M. (1990) ViewPoint Oriented Software Development, Proc. Conf Le Génie Logiciel et ses Applications, Toulouse, p 337-351 Franckson M., Peugeot C. (1991) Specification of the Object and Process Modeling Language ESF Report n° D122-OPML-1. 0. Gibbs W. (1994) Software's chronic crisis, Scientific American, Gnaho C. , Larcher F. (1999) A Goal Driven Approach to Web Navigation Modeling, Proc. Intemational Workshop on Internet Data Management, Florence, Italy. Gotel O., Finkelstein A. (1994) Modelling the Contribution Structure Underlying Requirements, in Proc. First Int. Workshop on Requirements Engineering : Foundation of Software Quality, Utrech, Netherlands. Humphrey W.S., Kellner M.I.(1989) Software Process Modeling: Principles of Entity Process Models, Proc. 11th Int. Conf. on Software Engineering. TEEE-830 (1984) Guide to Software Requirements Specification, ANSI/IEEE Std 830. Jacobson I. (1995) The use case construct in object-oriented software Engineering. In Scenario-based design: envisioning work and technology in system development, John M. Carroll (ed.), John Wiley, 309-336, Johnson J. (1995) Chaos : the Dollar Drain of IT project Failures. Application Development Trends, pp.41-47. Kardasis P., Loucopoulos P. (1998) Aligning Legacy Information Systems to Business Processes. Proceedings of the 10th Conference on Advanced Information Systems Engineering, CAiSE98. Pisa, Italy. Loucopoulos, P., Kavakli, V., Prekas, N., Dimitromanolaki, I. Yilmazturk,C., Rolland, C., Grosz, G., Nurcan, S., Beis, D., and Vgontzas, G. (1998) The ELEKTRA project : Enterprise Knowledge Modelling for change in the distribution unit of Public Power Corporation, 2nd IMACS International, Conference on Circuits, Systems and Computers (IMACS-CSC98), Athens, Greece, pp. 352-357. Lubars M., Potts C., Richer C. (1993) A review of the state of the practice in requirements modeling, Proc. IEEE Symp. Requirements Engineering, San Diego. Martin J., Odell J. (1992) Object Oriented Analysis &Design, Prentice Hall, Englewoods Cliffs, NIO7632. McGraw K., Harbison K. (1997)User Centered Requirements, The Scenario-Based Engineering Process. Lawrence Erlbaum Associates Publishers. Mylopoulos J., Chung, L., Nixon, B. (1992) Representing and using non functional requirements: a process-oriented approach, IEEE Trans. Software Eng. Vol 18, N 6. Olle, T.W., Hagelstein, J., MacDonald, LG. Rolland, C., Sol, H.G., Van Assche, F.J.M., Verrijn-Stuart, A.A. (1988) Information Systems Methodologies: A Framework for Understanding, Addison-Wesley. Pohl, K. (1994) The Three Dimensions of Requirements Engineering: a framework and its application, Information Systems Vol 19, N 3, pp 243-258. Pohl K. (1996) Process Centered Requirements Engineering, J. Wiley and Sons Ltd. Potts C. (1989) A Generic Model for Representing Design Methods, Proc. 11th Int. Conf. on Software Engineering. 18Potts C. (1997) Fitness for use : the system quality that matters most. Proceedings of the Third Intemational Workshop on Requirements Engineering: Foundations of Software Quality REFSQ'97 , Barcelona, pp. 15-28. Prakash N. (1997) Towards A Formal Definition of Methods, The Requirements Engineering Journal, 23 -50. Ramesh B,(1993a) A Model of Requirements Traceability for Systems Development, Technical Report, Naval Postgraduate School, Monterey, CA. Ramesh, B., Edwards, M. (1993b) Issues in the Development of a Requirements Traceability ‘model, Proc. IEEE Symp. on Requirements Engineering, IEEE Computer Society Press, San Diego, Ca. Ramesh, B., Powers T., Stubbs C. and Edwards, M. (1995) Implementing Requirements Traceability : A Case Study, in Proceedings of the 2nd Symposium on Requirements Engineering (RE'95), pp89-95, UK. Rolland C., Cauvet C. (1991) ALECST : An Expert System for Requirements Engineering, Proc. 3th Int. Conf. on Advanced Information Systems Engineering, Springer Verlag. Rolland C., Souveyet C., Moreno M. (1995) An Approach for Defining Ways-of-Working , Information Systems Journal, Vol. 20, No 4, pp337-359. Rolland C., Ben Achour C. (1997) Guiding the construction of textual use case specifications. Data & Knowledge Engineering Journal Vol. 25 N° 1, pp. 125-160. Rolland C., Souveyet C Ben Achour., C. (1998a) Guiding Goal Modelling using Scenarios, IEEE Transactions on Software Engineering, Special Issue on Scenario Management, Vol. 24, No. 12, 1055- 1071 Rolland C., Ben Achour C., Cauvet C., Ralyté J., Sutcliffe A., Maiden N.A.M., Jarke M., Haumer P., Pohl K., Dubois E., Heymans P. (1998b) A Proposal for a Scenario Classification Framework. Requirements Engineering Journal, Vol; 3, No. 1, pp. 23-47. Royce W. W. (1970) Managing the Development of Large Software Systems, Proc. IEEE WESCON 08. Rumbaugh J., Blaha M., Premerlani W., Eddy F., and Lorensen W.(1991) Object-oriented ‘modelling and design. Prentice Hall. The Standish Group (1995) Chaos. Standish Group _Internal_Report, http://www standishgroup.com/chaos html, Van Lamsweerde A., Dairmont R., Massonet P. (1995) Goal Directed Elaboration of Requirements for a Meeting Scheduler : Problems and Lessons Learnt, in Proc. Of RE'95 - 2nd Int. Symp. On Requirements Engineering, York, pp 194 -204 Wynekoop J. D., Russo N. L. (1993) System Development methodologies: unanswered questions and the research-practice gap,, Proc. of 14th ICIS (eds. J. I. DeGross, R. P. Bostrom, D. Robey) pp. 181-190. Yourdon E.E (1989) Modem structured analysis, Prentice Hall, 1989. Yu E. S. K, Mylopoulos J. (1994) Understanding Why in Software Process, Modelling, Analysis, and Design, Proc. of the 16th International Conference on Software Engineering Sorrento (Italy). Yu E. S. K, Mylopoulos J. (1994) From ER to AR modelling strategic Actor Relationships for Business Process Reengineering. Proc. of the 13th Intemational Conference on the Entity-Relationship Approach, Manchester, UK. Yu E. 8. K, Mylopoulos J. (1994) Towards Modelling Strategic Actor Relationships for Information Systems Development- with Examples from Business Process Reengineering . Proc. of the 4th Workshop on Information Technologies and Systems, Vancouver, Canada. Zelnick N. (1998) Nifty Technology and Nonconformance : The Web in Crisis, Computer, PPIIS-119 19VLSI Design Education for Undergraduate and Graduate Students Yoshikazu Miyanaga Hokkaido University, Graduate School of Engineering, Sapporo, 060-8628 Japan miyanaga@ice.eng hokudai.ac.jp Abstract In this report, two recent: VLSI education and resoarch centers in Japan are introduced. Both of them have been founded for the establishment of Japanese high VLSI education and research in an university. One is VLSI Design Education Center (VDEC). It mainly supports LSI design education in “universities where university staffs operate this center. Other is Semiconductor Technology Academic Research Center (STARC). It mainly supports LSI research and development in universities where ‘many researchers of VLSI fabrication companies have the role of research promotion. ‘This report Introduces the basic roles of both centers. In addition, two aetual projects which the author's group. has started and advanced are also introduced. 1 Introduction ‘The essential role of VLSI has been recognized in all of products. If any product would be developed with a small size, alow cost, a low power consumption, a high speed and a large number of productions, the LSI design must be considered and the main part of the product should be replaced into VLSI. In recent, any of Key technologies should be considered at the point of very low cost production, In other words, the VLSI design and fabrication seems to support all of recent Key technologies. ‘The rate of an added value on VLSI in any product becomes getting increase as shown in Fig.l (1). Even in 1998, the 26% functions of a product have been designed by using VLSIs. In the near future, almost half number of functions should be implemented in VLSIs. ‘When we consider the world market of VISI, its actual market size becomes quite large compared with several years ago (1). In addition, the fields of computer and communication are getting larger than others as shown in Fig.2, It is also considered that the technologies used in computer and communication are getting to be applied into home electronics which is called intelligent home electronics. VLSI Rate of an added value in a product | VA "Yoo 1993 19 Figure 1: Rate of an added value to VLSI in a product (BIAJ.NRI 1999) Although the circumstances on VLSI design and fabrication become quite important and every Japanese company have spent quite large amount of capitals to the development of LSI technology, 2122 1906 645 ition ae on 1 mComputer 158 Communication 150% Home Electronic Py Others 1999 $124 Bition Figure 2: VLSI World Market (WSTS 1998) the LSI education and research in Japanese universities were not made brisk until several years ago. The primary factor of such condition might be taken on the distance between the hardware design education in universities and the VLSI research in companies. Based on such background, two important centers ‘were organized in Japan. One of them is VLSI Design Education Center (VDEC). The center is located in University of Tokyo 1 mainly supports LSI design education in universities where university staffs operate the center. Other Semiconductor Technology Academie Research Center (STARC). This conter has been organized by 11 Japanese LSI fabrication companies. It mainly supports LSI research and development in universities ‘where many researchers of VLSI fabrication company have the role of research promotion. ‘This report introduces both centers briefly. In addition, two actual projets which the author's group has advanced are also introduced. 2 VLSI Design Education Center (VDEC) in Japanese Univer- sities ‘The VDEC started from May 1996 [2,3]. The VDEC center is located in University of Tokyo and other 9 universities including Hokkaido university support it as VDEC sub-centers shown in Fig.3. ‘The main purpose of VDEC is the good support of the VLSI design education in Japanese universities and colleges. ‘As shown in Fig.4, many professors in universities, e.f., 399 in 1998, cooperate with this center. The number of users are also getting increase, For this purpose, the center is usually providing any kind of knowledge on VLSI design and the actual fabrication of LSI chips with several Japanese companies, ‘There are two main services. One of them is the supply of VLSI CAD systems and some seminars on ‘VLSI design. As to the CAD tools, we have prepared the following systems: 1, Cadence Design System ‘This is used for Verilog HDL logic synthesis, its simulation and analog design 2. Synopsys Design System ‘This is used for VHDL logic design and its simulation. 3. Appollo System ‘This is used for automatic placement and routing ‘These CAD systems can be used from any universities and colleges in Japan if they would like to use CADs. These systems are implemented in the center and the subcenters. The number of users are shown in Fig. In addition to the CAD systems, several text books have been prepared and published. These books are written as self-content and an undergraduate student may read and study VLSI design with them by himself. We also provide several tutorial courses without or with some supported CAD venders ‘two times in every year.Figure 3: VDEC center in University of Tokyo and Sub-centers in which Hokkaido university is depicted as a black box. (VDEC 1998) Figure 4: Number of Professors in VDEC (VDEC 1998) AAs the other main service, the center keeps the connection between the universities and the LSI fabrication companies. The designed LSI can be fablicated as a real chip with quite low cost. This must bbe important circumstances for university students. Any student: who designs his/her chips can obtain actual chips and test them in the center and the subcenters. 3 An Example of LSI Design in VDEC 3.1 Structured Recurrent Neural Network In this section, a VLSI design and its error analysis for a structured recurrent neural network are intro- duced. The structured recurrent neural network has been proposed by Dr.Li et. al. [4]-[7]. Its method is well developed for the suitable concurrent LSI system. Accordingly the whole architecture can be designed as a LSI system. ‘A typical recurrent neural network is a full-connected model which is not easily implemented with a VLSI system (4), [5]. Accordingly, we have designed a new network structure. It is called as a structured recurrent neural network. The network topology is redesigned for a VLSI system and thus the function is fairly limited. However, as quite important characteristics, the dynamic behavior and feedback mechanism are still implemented. As the main modules of the network consists of two simple nodes and each node hhas fully connected to other and itself, the total connections are considerably reduced and its structure must be suitable for VLSI design. For this method, we have really developed a part of the network in a VLSI system. In the design of VLSI system, the parallel pipeline mechanism is introduced to get the high processing speed. In addition, binary look-ahead carry adder circuit (BLCA) for an adder/subtracter and a carry save adder circuit (CSA) for a multiplier are also redesigned to get high speed arithmetic circuits, As one of important 2324 Figure 6: Number of Chips (VDEC 1999) issues, the system should be designed as small as possible. ‘This report in particular introduces the theoretical error analysis of fixed-point operation and shows experiments to obtain the optimum word Jength of the SRNN VLSI system [8]. The designed VLSI chip has been fabricated using 0.35 ym CMOS technology of Hitachi. 3.2. Algorithm of SENN 7 shows the structure of SRN. The basic module is depicted in Fig.8. In Fig.8, two conventional neural nodes are used and fully connected. By using this basic unit of Fig, any regular limit circle can be represented. In addition to these basic units, the center node in Fig.7 has bi-directional connection to each node in every basic unit. Compared with the fully connected recurrent neural network, its behavior ‘may be limited. However, some chaotic properties can be still realized by using this SRNN. ‘At the point of LSI design, the full-connections to many nodes are not suitable for a chip design. In ‘other words, if some restrictions can be introduced for the proper design of VLSI, the area of connections in VLSI may be considerably reduced. The above SRNN consists of local full-connections and a few ‘number of connections in proportion to the number of basic units, In the basic unit, the dynamics of the node is defined as ried) = ae) + Fe +5 a) where i = 1,2 and , we = Sea(o) @ se) = eS @ ‘The variables 24(t) and y(t) are defined as the intra-cellular membrane potential and the firing rate (output) of the -th neuron, respectively. The value 7 is @ decay time constant and s, is the external inputs. The value wi denotes the weight connecting the output of the j-th neuron to the input of theInput Output Figure 7: Block Diagram of SRNN Figure 8: Block Diagram of SRNN ith neuron. The value Nis a number of neuron units in the basic units. The S(2) is called a biased sigmoid function, In the output layer, the dynamics of neuron is defined as drous(t) dt —touelt) +3 wouess(t) @) where y;(¢) is the neuron output of the j-th basic unit, All outputs are feedback to the corresponding input neuron. 3.3. Error Analysis ‘This subsection intzoduces the theoretical error analysis of fixed-point operation and shows some exper- {ments to estimate the optimum word length of data for the SRNN VLSI system, First a round off error is analyzed to determine the optimum word length of data. ‘This analysis is quite important since the total gate number is almost in proportion to the determined word length. Accordingly, any superfluous bit length should be deleted. For the effects of round off error on a multiplication, the following error variance is referred (0) am sae ©) where o? denotes the error variance of each fized-point multiplication. The value n is the word length of decimal portion The theoretical round off error variance of SRNN is shown in Fig.9. In Fig.9, X-axis represents the word length of decimal portion (n) and Y-axis represents the error variance. When the word length of decimal is set around 10, the error vatiance gets a tendency of approaching to zero. 25Figure 9: Round off error variance of a SRN System 3.4 Design of VLSI System ‘When the system is designed in all digital circuits, differe function should be approximated suitable for digital calculation. ‘The differential equation of an unit is rewritten as follows: ‘equations and a non-linear sigmoidal aint) = (1-Apain) + Py (ay(ny) + 4 © ‘The non-linear sigmoidal function is quite simply approximated as 10 2>20 wlz) = ) 052 20>2>-20 o U-10 2<-20 ‘The basic unit is constructed by using 5 parts, ie., an adder/subtrocter, a multiplier, a sigmoid ‘operator, a controller and register blocks. In order to minimize every circuit scale and path delay, a suitable arithmetic logic circuit is redesigned ‘An adder/subtractor is built by a binary look-ahead carry adder (BLCA) circuit. A multiplier is ade oft carry save adder (CSA) circuit. Since the slope is selected as 0.5, the sigmoid function ean be designed by a I-bitright-shift unit. After those arithmetic logic circuits are used, the evaluation of the cell-area and the path delay of every circuit is described in Fig. 10. onan] Todas @ ua) Soot [aang | ie | rao Cex | acwtsa [an | sup) ed [ve | ix sioae_[ aout | ent [ence (vitae | — es) [oust Me Tacs [a8 [na Tea [eet | — 1s [rae sen [fete [68 [wren Figure 10: Evaluation of Total Area and Delay ‘A part of SRNN has been designed in @ VLSI system through Hokkaido VDEC. To describe e hardware design, the verlog, hardware description language (Verilog HDL) has been used. In addtion, the Verilog- XL digital logic simulator, the Design Analyzer and Placement/Routing tools have been used to excite 26the logic simulation, the logic synthesis, optimization and layout design of the system. By using these tools, the LSI chip has been designed with 0.35 ym CMOS technology. Fig.11 shows the whole upside view of the fabricated chip. Figure 11: Upside View of a Designed SRNN chip 4 Semiconductor Technology Academic Research Center (STARC) ‘The Semiconductor Technology Academie Research Center was developed by 11 Japanese companies who have quite important role of VLSI vendors in Japan (10},(11}. The purpose of this center is mainly the support to the cooperation and associations between universities and companies. It started just 3 years ago. The research cooperation among Japanese universities and companies have apparently long history. However, few actual products and technologies have been developed in the world market from such cooperation unfortunately. In this center, many important technologies in universities have been explored and developed as the new products and technologies. 1 Fig.12, the process of this center is summarized. Fig.12 depicts the followings 1, Support from companies The 11 companies support the capitals to this center. In addition, some visiting research staffs are usually elected and then they support the specific project of the university research group. 2, Cooperation with universities ‘The university research group which is interested in the project with STARC first. proposes its research plan. This plan should be concluded in 3 - 5 years. Once this plan is selected by STARC, the new project is going to start with a research staff of STARC and some company visiting researchers. 3. Cooperation with companies ‘The STARC also proposes his research plan to some companies which are called clients. If the proposed plan is accepted, STARC is going to start the new project with some university groups. 4. Supply to others All of products and technologies in STARC will be opened to others. If some supports will be required in case of an actual system, the associated university stafls and STARC research stats wil help its development AAs shown in Fig.13, only 5 permanent research staffs are employed in STARC. However, many re- searchers from companies are cooperated with this center. As a unique mechanism in Japan, they have 1 mission to fully help the project. In other words, the project is not promoted by only university staffs but also supported by both of university staffs and visiting company researchers. Both of them have the same responsiblity on the project. In addition, the project is usually evaluated by a STARC reserach staff in every 3 mounth. ‘The execution of the budget is dynamically reflected by this evaluation. 2728 () @) Q 0} Figure 12: Mechanism on STARC 89 9 Mt Research Suff “o m4 Visiting suff 8 Project ao} | 1996 1907 1998 1999 Figure 13: Number of STARC Projects and Research Stalls 5 An Example of LSI Design in STARC 5.1 Data Classification and Labeling ‘This section introduces an analog LSI architecture for data clustering and labeling. ‘The classification and the labeling of data are an essential method for media processing, e., speech recognition, image recognition and so on. These methods usually require a large calculation cost of linear/nonlinear data processing. Accordingly if this method is designed as an efficient architecture, high speed data recognition and a small scaled processing system can be obtained. This section shows the analog LSI system in which Euclidean distance is fist realized for date clustering and in addition in which a simple perception net is realized by a neuron MOS. Since the structure of designed circuits is quite simple, the total number of transistor is considerably reduced. As an exmaple of data clustering, this report shows the results of HSPICE simulation for Japanese vowel recognition. 5.2 Mechanism of Clustering system “The system consists of two clustering layers as shown in Fig. 14. ‘The first layers a selfonganiaed clustering layer andthe second layer i labeling lye. The network weights and all node information ofboth layers have determined by the taining proces. In the proposed aystem, i is assumed that al trainings are accomplished and all of information are prior implemented inthis syst, In {12}{15), the ayete does not use Eucidean distance but Standard Euclidean datance. However we consider that Euclidean distance may successflly bulld the clustering ayatem in case of the much increase of odes The basic calculations are given at follows: Consider there are m cluster nodes and each node has the p-th dimensional node pattern vector, i.e., 24(i = 1,2,---,m). Each node calculates the distance, ie, D4(i= 1,2, betwoun an input vector, be y= (ys¥or"*-sHp) and a pattern vector, Le, 24=Label | Label2 ++ Label Output Figure 14: Structure of the clustering system (cua,2iay + -e2ip). To realize non-linear property, the threshold value Dry is set up as follows: 5) 8) Ls 1-(D/Drw? Dy < Din {a OP Som ® ‘There are n nodes in the scond layer and each node has the m-th dimensional weight vectors, i, we (Wa thay") (t= 1,2,--5,8). The output of the second layer is neaSs ao ‘The output label is a vector, i., 2= (21,22,°+",2n), which indicates the signs of Re 0 R<0 a(t no en 5.3 Circuit of Clastering/Labeling System In order to design the clustering/labeling system which uses the Euclidean distance, the neuron MOSFET (oMOS ) structure is considered to be useful [16]. Let us consider MOS transistors connected with n ‘capacitors and a switch shown in Fig.15. First, Vi, Va,-++ and Vq add to each nodes and Va also add to MOS gate (pre-charge). When the switch is opened, the MOS gate becomes floating. After that, the voltages of the left nodes are changed to Vj, Vj,--» and VJ. In this case, the floating node potential becomes +Va (12) ‘This circuit is used for the subtraction in the first layer and the weighting calculation in second layer. ‘The agent network is designed in all analog circuits. In the frst layer of this network, the basic, calculations of (8) and (9) are applied [17]. In addition, the node of the second layer executes the calculation of (10). Fig.J6 shows the analog circuit which realizes Eqs (8) and (9). In addition, Fig.15 shows the analog cireuit which realizes Eq, (10). 5.4 Hardware Architecture of Speech Recognition System By using the analog agent network, we have designed a speech recognition system. The proposed speech recognition system consists of mainly three stages shown in Fig.17. In the first and the second stages, 2930 Yue Voge Ve 15: vMOS few“) ee [ee “ Figure 16: Analog Circuit of Distance Measure this system calculates the characteristics of observed speech. After them, the data clustering and labeling should be executed. In the third stage, ie., the data clustering/labeling, the above analog agent network is epplied. ‘As to the first stage, an ARMA lattice filter is used. This filter is described in Fig.18, As to the ARMA lattice modeling, its architecture has been designed in Fig.19. Algorithm A calculates the prediction error and the estimated input. This is executed in mainly parallel processing. Algorithm B calculates the backword prediction errors which are used in the parameter estimation of the next time instance. This calculation can be executed in fully pipelined processing. The block of BB in Fig.19 is an elementary block. In the designed system, the EB block is implemented in digital ALU and registers, Using the designed EB repeatedly, ARMA lattice parameters are calculated. From the ARMA parameters, 32 points spectrum data can be calculated in parallel processing. These spectra data are temporally stored in registers and arranged to be an input vector to an agent cluster- ing labeling network. 5.5 Simulations The SPICE simulations of the clustering/labeling system used for the Japanese vowel recognition are shown in this section, The speech data was analyzed by ARMA spectrum estimation. The 233 spectrum envelope vectors were prepared from vowels spoken by a woman. Repeating 2000 times of the learning process in the software system, all capacitor values were determined. Note that they were round offSpeech Waveforin Adaptive Analysis of ARMA Lattice Filter Time-varying ARMA Lattice parameters Evaluatios ‘Speech Spectrum Envelopes Speech Spectrum Envelope Two Layer Clastering’Labeling Nework Recognition Results Figure 17: Block Diagram of Speech Recognition System 256 steps. Input vector is 30 dimensions, the node number of the first layer is 20, and the output label number is 5. One cycle set 250ns. The total devices of the simulation circuit, are only 1570 MOSs and 3931 capacitors. First, when all of learning vector data were fed into the circuits one by one, 3 vector data were 1misrecognized in this proposed circuits. Fig.20 shows the voltage of a comparison floating gate and /a/, /i/, Jf floating gate where the input vectors are the spectrum data of /a/, /i/ and /u/. ‘Next, the Monte Carlo simulation was held. Each nMOSs and pMOSs were changed the parameter of the threshold voltage Vj, and transconductance KP. The value Vix changes 0.05V and KP change 5% in o = 1 level as Gauss distribution. The result of the simulation is shown in Fig.21. All of 30 times simulation provided correct behaves in the clustering/Iabeling system. 6 Conclusions In this report, two VLSI education and research centers in Japan are introduced. Both of them have been founded for the establishment of Japanese high VLSI education and research in an university. One is VLSI Design Education Center (VDEC). Other is Semiconductor Technology Academic Research Center (STARC), In VDEC, the center is providing any kind of knowledge on VLSI design and supporting the actual fabrication of LSI chips with several Japanese companies. In particular, the CAD systems which are ‘introduced in many VLSI design companies can be usually used in the center and the sub-centers. In addition, several design rules are actually opened to VDEC users for the VLSI fabrications. A student can design a chip and finally get the real chip. As an example, the design of the recurrent neural network which is restricted at the point of structure is introduced. In the chip, a part of the proposed network is 31only realized, However, since the design of a small scaled system and the error analysis of total calculation are considered, it can be expected to get the high performance RNN VLSI system with small size. The purpose of STARC is mainly the support to the cooperation and associations between Universities and companies. This center is now trying to develop the new products and technologies with Japanese universities. As one of examples, this report shows the designed architecture of a speech recognition system. In particular, it is shown that the excellent analog circuits can realize the data clustering and labeling. Figure 18: Example of ARMA Lattice Model for Spesch Analysis EB Block f T. Figure 19: Pipeline Architecture of ARMA Lattice Modeling, 7 ACKNOWLEDGEMENT The author would like to thank Research and Development Headquarters, Yamatake Corporation for fitful discussions. The VLSI chip in this study has been fabricated in the chip fabrication program of VLSI Design Education Center (VDEC), the University of Tokyo with the collaboration by Hitachi Ltd., and Dai Nippon Printing Corporation. This study is also supported in parts by Semiconductor Technology ‘Academic Research Center (STARC) and Japan Science and Technology Corporation (JST) in Japan. 32Voltage of Floating Nodes (50m Vidiv) Tine (S0nsidiv] Figure 20: Voltage of floating nodes ‘Ouput Voltage [2V/dv) L z rl “Time [1251010] Figure 21: Monte Carlo simulation References [1] Hajime Sasaki, "Information Society and Semiconductor Technology”, 2000 Workshop on Semicon- ductor Technology for the 21th Century, vol-1, pp.1-3 - 1-25 (June 2000) [2] "VDEC Annual Report 1997", VDEC in Japan, printed in Japanese (1997). [3] "VDEC Pamphlet 1999", VDEC in Japan, printed in Japanese (1999). [4] Y.Li, Y.Miyanaga and K-Tochinai: "Memorizing and regenerating spatiotemporal patterns with structured recurrent neural network” , Trans. IEICE, Vol.J79-A, No.1,pp29-37, 1996. (5] K.Doya and $.Yoshizawa: "Adaptive Neural Learning”, Neural Networks, Vol.2, pp.375-385, 1989. (6] HL.Li, ¥.Miyanaga and K-Tochinai: "A VLSI Design and its Error Analysis for Structured Recurrent Neural Network”, IEEE International Symposium on Intelligent Signal Processing and Communica- tions 1999, vol.1, pp.65-68, (December 1999), stor using Continuous time Back Propagation [7] HLLi, Y.Miyanaga and K‘Tochinai: "A Hardware Design for Structured Recurrent Neural Network”, Proceedings of International Technical Conference on Circuits, System, Communications and Com- puters, vol, pp.581-584, 33{8] J.Ohkubo, Y.Miyanaga, K:Tochinai: "A CPLD Design of a Self-organizing System for Data Clus- tering”, IBEE Processdings of International Symposium on Circuits and Systems, 1998. {9] M.Kawamata and T-Higuchi, "Minimum Round-off Noise Realizations of Separable Denominator ‘Two Dimensional Digital Filters”, Trans. IEICE, Vol.J68-A, No.11,pp.1200-1208, 1985, {10] Y.Miyanaga, "Robust Spooch Recognition Techniques and its VLSI Design”, STARC Symposium 1999, vol.1, pp.107 - 116 (September 1990). [11] "1999 STARC News No.2", STARC in Japan, printed in Japanese (1999). (12) Ruslam, ¥.Miyanaga, K.Tochinai, ”Multi-clustering Network for Data Classification System”, Trans, EICE on Fundamentals of Electronics, Communications and Computer Science, E80-A, 9, pp.1647- 1054, September 1997. jon techniques and its VLSI design”, Proc. of Stare [13] Yoshikazu Miyanaga, "Robust speech recogn ‘Symposium 99, pp.107-116, September 1999, (14) Yoshikazu Miyanaga, "An adaptive speech recognition system and its VLSI design”, Proc. of 1999 International Symposium on Nonlinear Theory and its Applications, pp.883-886, Dee. 1990, (15] H.Nagahams, Y.Miyanaga and N.Ohtsuki, ” An adaptive speech analysis for speech recognition sys tem”, Proc. of IEEE ISPACS 1999, pp.745-748, Dec. 1999, [16] M-Konda, T:Shibata and T.Ohmi: *Neuron-MOS Correlator based on Manhattan distance compu tation for event recognition hardware”, IEEE Intemational Symposium on Circuits and Systems 4, p-217-220, 1996. [17] Ugur Gitingirojglu, Devrim Y.Aksin, "A 4-Transistor Euclidean Distance Cell for Analog Classifiers", IBEE Processdings of Intemational Symposium on Circuits and Systems, 1998.DESIGNING 3G APPLICATIONS FOR SMARTPHONES AND WIRELESS DATA TERMINALS Gareth Loudon Cyberlab Singapore, Ericsson Research 510 Thomson Road, #18-00 SLF Building, Singapore 298135 ABSTRACT Excellent cellular bandwidth, instant access and fast response times finally challenge the world of fixed Internet. The future world of cellular communications offers space not only for applications on smartphones and PDAs but also terminals as large as books. But how should you go about developing applications for these devices. Cyberlab Singapore approaches possible applications and devices for future cellular networks by designing for particular cultural groups and environments. Examples will be given on work done in the area of applications and devices for teenagers, Chinese business users, as well as for particular environments such as hospitals, cafes etc.. INTRODUCTION ‘The term Mobile Internet has been well publicized recently. Figure 1 shows the projected subscriber growth over the next few years highlighting the tremendous increase in Mobile Internet usage. 1380 1200 1080 900 750 Fixed internet (Millions) 450 300 Mobile Internet 180 ° 1996 1997 1998 1999 2000 2001 2002 2003 2004 2006 Figure 1. Subscriber growth for fixed and mobile usage (including Internet).But what does the Mobile Internet really mean for the user in the future 3" Generation (3G) networks, and what are the challenges and opportunities that lie ahead for telecommunication companies, service providers and third party developers? Before this can be discussed in more detail it is important to be aware of some key benefits that come from the next generation of wireless communication technologies. There are several new generations of wireless networks in the pipeline that evolve from the current 2" Generation GSM technology and expand the capabilities of the network, The new networks move to a packed-based mode of communication, allow high peak bandwidth, direct IP access, short response times and the capability for users to stay permanently on-line. With these changes in the wireless network infrastructure in the future, many more compelling applications are possible. Below is a short description of the network evolution planned. HSCSDIGPRS: The first step in the GSM evolution is known as 2G+. This includes high user bit-rates through multi-slot High Speed Circuit Switched Data (HSCSD) and General Packet Radio Services (GPRS), which will offer a data rate up to 115kbit/s. As a packet-based technique, GPRS will enhance GSM data services significantly — especially for bursty Intemeviintranet traffic — and make optimal use of available radio spectrum and channels. With the new GPRS nodes, direct IP access from the radio access network will be provided. GPRS will allow many users to share the same channel and allow users to stay virtually "on line" all the time: radio resources will only be used when data is actually being transmitted or received. Call set-up will be almost instantaneous and users will be charged on the basis of actual data transmitted, rather than connection time. EDGE: The second step for evolving GSM is the implementation of Enhanced Data rates for GSM Evolution (EDGE), known as 2G++. EDGE will allow GSM operators to use existing, GSM radio bands to offer wireless multimedia IP-based services and applications at speeds of 384kbit/s with a bit-rate of 48kbit/s per timeslot and, under good radio conditions, up to 69.2kbit/s per timeslot. EDGE will allow the full advantages of GPRS to be explored, with fast set-up of connections and higher bandwidth than traditional GSM. The combination of EDGE and GPRS will result in much improved utilization of the radio network. EDGE will be beneficial to existing operators and when small portions of spectrum are made available in current bands. Channels with EDGE functionality will be able to operate in either existing GSM or EDGE mode, allowing the coexistence and seamless step-by-step introduction of EDGE. WCDMA: The third step in the GSM evolution is the new WCDMA (Wideband Code Division Multiple Access) air interface standard on the GSM core network, planned to provide multimedia access at up to 2Mbit/s. This capability will be particularly beneficial when large portions of new spectrum are made available, e.g. in the new paired 2GHz bands in Europe and Asia. Further benefits of WCDMA include reduced fading — a result of the improved frequency diversity effects of the wider SMHz channel bandwidth. This is one of the technologies that will be used for the 3" Generation systems. Figure 2 highlights the evolution in time from GSM to WCDMA and the benefits in bandwidth and services. 36Bits High ey CSE ua SCO quality ere pa Figure 2. The Evolution of GSM and possibility for new services. FUTURE CHALLENG 'S IN THE MOBILE INTERNET WORLD. Let's revisit the question of the challenges that lie ahead for companies. One of the main challenges for companies is to come up with exciting and appealing applications that run on this new wireless communication infrastructure. This may seem straightforward, but there are some specific challenges for Mobile Internet access. One of the first challenges is to handle the problems that arise from the difference between mobile channel characteristics and fixed line characteristics. Another key problem is the mobile phones themselves. Mobile phones (or other mobile devices) are traditionally physically small in size. This means that a mobile device will probably have a small screen size and in many cases no keyboard If this is the case, how is it possible to interact with the device and navigate through remote Internet sites, which contain a large amount of content, to find the information you need? After all, most desktop computers today rely on the fact that they have a large screen to display content, a keyboard to allow text entry and a mouse to provide further interaction capabilities. To make matters even worse, most mobile phones only have a numeric keypad and a small number of cursor keys! One solution is to allow other modes of interaction, such as voice and/or pen based interaction on a mobile device. These new modes of interaction can provide a solution to the problem of small screen sizes, the lack of a keyboard and the lack of a mouse. Pen based interaction will mean that the small screens will be touch sensitive, and therefore take over the role of a mouse. The pen could also be used for text entry via handwriting, if handwriting recognition technology is incorporated into the mobile phone. It could also be used for drawing pictures and scribbling down handwritten notes for either keeping yourself orsending to your friends or colleagues. Voice based interaction could be used for the activation and control of the functions in your mobile phone, like making a phone call (as is possible to some extent today) but also for searching an Internet site, whether it be for MP3 songs, stock information, companies on the yellow pages or whatever. It could also be used for text entry. Whether the speech recognition technology resides on the phone or on a remote server may well depend on the type of application, Figure 3 gives an example of one of Ericsson’s future concept phones that integrates pen and voice interaction for control and navigation. Figure 3. A concept phone by Ericsson that highlights multi-modal interaction and Bluetooth capabilities, Figure 4. A concept terminal by Cyberlab Singapore, Ericsson Research, with a large 10” screen (the DelphiPad) 38Figure 4 gives another example of a concept terminal, developed at Cyberlab Singapore, Ericsson Research (known as the DelphiPad), that has a large 10” screen and is designed predominantly for mobile data communication. So the important messages are that the mode of interaction and navigation, as well as the useful applications, for the Mobile Internet may well be different from that of the current Internet. Therefore if this is true, how should you go about designing and developing useful and easy-to-use applications for these mobile devices. This is the question Cyberlab Singapore tries to answer. DESIGNING 3®° GENERATION MOBILE APPLICATIONS Cyberlab Singapore approaches possible applications and devices for future cellular networks by designing for particular cultural groups and environments. By choosing and studying particular cultural groups (eg. Teenagers, or Chinese business users) or particular environments (eg. Hotels, cafes, or hospitals) it is possible to understand the key problems, challenges and unmet needs of users. Using this information together with an investigation of what is actually achievable technically, it is possible to create innovative mobile applications. The following section gives an example of how this approach has been used to design a ‘mobile application for teenagers. DESIGNING A MOBILE APPLICATION FOR TEENAGERS At Cyberlab Singapore, we divide the application design into four main stages: Hypothesis, User Studies, Interaction Design and finally Prototype Creation. For the teenagers we started with the following list of hypotheses that we thought were interesting and relevant with regard to the next generation wireless technology and the Internet revolution + They want to feel part of a group and share their experiences and emotions with friends. + They always want to be connected to their friends. «They have time on their hands. + They want to express their individuality and influence and impress friends. They want to be cool! + They are very mobile and want to be up-to-date on the latest information (be it fashion, music, movies, gossip etc.) anywhere, anytime. + They are excited, interested and comfortable with new multimedia, Internet and mobile phone technology. We then came up with a concept that makes the most of these traits and is also exciting and fun to use, and hopefully appeals to a very large consumer group, on a global scale. Some of the key ideas in the concept were: © People centric, Real-time jon of current activities or emotions of friends. Simple access to multimedia content on the Internet and on other peripheral devices, Sharing of multimedia content amongst friends. Group communication and collaboration through multimedia and voice. ‘Natural language and multi-modal (eg. Speech and handwriting) communication with the device as well as friends. 39These ideas were then translated into an application context on a mobile device, by creating a QuickTime Cartoon movie. Some of the main modules in the application are listed below. © A real-time view of the state of close friends, in terms of their emotion, activity or comments. * A place to express your own personal emotions, activities and comments. ‘* A scribble pad to jot down notes (with a pen) for sending to friends or sharing A place to store the multimedia content or notes from friends that you want to keep. A place to communicate with your agent to retrieve remote information. After the hypothesis had been made about a possible mobile application for teenagers, the next stage was focused on user studies. Studies were conducted in several countries around the world. From the user studies we gained a better understanding of teenagers" lifestyles and attitudes as well as getting direct feedback on the application concept. The results of the user studies were used as direct input into the third stage, Interaction Design. The Interaction Design stage involved the creation of a clear application architecture using the analysis results from the user study stage, as well as from a technical feasibility study. The final stage was then to create the prototype application for larger scale user trials. Figure 5 highlights the stages taken in the development process. Figure 6 shows different screen shots of the application created after the prototype stage. Target Group Figure 5. The different stages in the application development process. 40weep ou ‘AD About My Meter ‘Anarcan Beaty ‘Argpkis Asker ‘Any Given Sindy ‘Are You Lonesome Tonight Balto Earth Beant Lg. cra ren) Figure 6. Screen shots of some of the different elements in the teenager application. The process allowed us to refine our initial hypothesis, emphasize some features more than others, add features that we had missed, and remove features that users disliked. More importantly, it enabled us to create an application that is more in line with the needs and liking of teenagers. OTHER APPLICATIONS AND DEVICES APPLICATIONS FOR CHINESE PROFESSIONALS We have applied the same application development process to another cultural group, Chinese professionals. The needs of Chinese professionals are probably different from the needs of teenagers in Europe say, both at the application level and at a language level Interaction with the mobile device is a specific area of interest for Chinese users, because of the large differences between the Chinese language, with thousands of ideographic characters, and European languages, with a small number of characters. Therefore our initial list of hypotheses of key concepts for Chinese professionals was quite different from the list of teenager hypotheses. One hypothesis was that a text entry solution using a keypad might be preferred for Europeans, while a text entry solution using voice or pen might be preferred for Chinese users. By using the same application development process described in the previous section you can find out whether these hypotheses are valid or not . Figure 7 shows how the same process has been applied for the Chinese users. 41Target Group User Studies Figure 7. The application development process for Chinese users. APPLICATIONS FOR LARGE SCREEN, MOBILE TERMINALS. The large size concept terminal, called the DelphiPad (shown in Figure 4), was created to explore application opportunities where a large screen, mobile device is necessary. Examples of where a large screen mobile device might be more appropriate than a very small mobile terminal could be in hospitals, hotels or even cafes. In a hospital for example, it could allow the Doctor, during ward rounds, to look at patient information, such as medical history, X-Rays ete. and to scribble notes of observations and diagnoses. This would not be as ea with a small mobile device or a laptop that is not suited to such mobile usage. Again the same development process would be used to create the application. In this case the user studies would be done with Doctors themselves to make sure that the application suited their needs. SUMMARY The future mobile networks and the emergence of the Mobile Internet will mean the availability of new services that have never been possible before, even on the traditional Internet. By the year 2002 you will see WCDMA networks (3 Generation systems) supporting very high bandwidth solutions. This means a total change in the way people will communicate, conduct business and keep in contact with family and friends. There are therefore great opportunities for the creation of new and exciting mobile applications. The race is on to find the killer applications and services for the new Mobile Internet. At Cyberlab Singapore we are trying to find the best solutions by designing mobile applications for particular cultural groups and environments and by involving users early on in the design process. 42E-BUSINESS IN SOUTH EAST ASIA Mark Gabbott Department of Marketing Monash University, Australia INTRODUCTION AND CONTEXT “It took radio over 30 years to reach 60 million people worldwide. It took television more than 15 years to reach 60 million people worldwide. I took just 36 months for the Internet to reach 90 million people worldwide.” It has become almost commonplace to assert that a new economic era is emerging in which the substance of business ideologies, practices and values will be transformed. The dominant view as expressed by Hepworth (1997), Knoke (1996) and Merry (1995) is that this new era will see developed economies become totally dominated by intangible knowledge based products and business assets such as intellectual capital, structural capital in the form of systems and customer relationships, will become the new economic drivers of market value. For economies while highly developed service sectors this shift in focus is a small step, but for those economies in S.E Asia where some are primarily manufacturing focused, this step is a long one. We can already see change emerging in the financial valuation of companies like Yahoo and Amazon.com who have market values of approximately $US 31 and 21 Billion respectively and yet provide nothing more than an infomediary service (Hagel 1997). Recently Qantas asserted that their most valuable company asset was their customer database, not staff, not planes, not routes just information. As the debate about the emerging information economy and the knowledge based enterprise emerges, there is a consistent view that at some fundamental level the information economy is driven by networks, and that it is the capacity for creating, using and leveraging value off information networks in particular which will be the key competitive advantage in this new information era (see for example Carr and Snyder 1997, Tetzeli 1992, McGaughey and Mason 1998, Berthon Pitt and Watson 1996, Bell 1997) Perhaps the most ubiquitous, most widely used and most often cited information network is the internet and its user friendly face, the world wide web. The apparent association between the internet and the information era has lead a number of authors to suggest that the ability to participate and capitalise upon the internet is a good indication of which businesses and which sectors will be capable of surviving in the new economic era (Negroponte 1997, Tapscott 1997). These commentators point to transformed business processes where the internet melds seamlessly into the organisational structure and provides the first signs of competitive market advantage for companies investing in pure information systems. (see Venkatraman and Henderson 1998). We also know that without the internet we would have been without some significant innovations in marketing practice such as ‘free’ information products, on line customer service and 24/7 transaction opportunities all bound up with a ‘cyber culture’. We also know that the internet is a challenge to accepted wisdoms, the real possibility of mass 43customisation, of one to one contact with customers and the end of competitive advantage based upon geographic location or bounded price information. Taken as a whole the internet is changing our perceptions of the marketplace, opening up new trading possibilities, defining new business models, and continues to mesmerise boardrooms across the world. We can summarise some of the key benefits of an e-business format which have been identified by participant companies: © The ability to trade across geographical borders and to access world markets at low cost and the reduction in investment hurdles associated with E-commerce allows small companies to compete with large established ones enhancing contestability © The creation of new distribution and sales channels coupled with the ability to re-align channel structures according to market demands and the broader economic environment ‘* Greater opportunities for developing customer relationship management techniques in large atomistic market environments including customer value analysis and data tracking © An ability to personalize and target a marketing message utilizing multi-media at very unit cost and provide key metric for advertisements, promotional offers and behaviour. © The ability to operate partnership marketing, linking websites to the sites of their business partners © Reducing the cost of business operating by streamlining customer, delivery, production and process management including paper systems. * A forcing of the new power dynamic where customers hold market power and not producers, where marketing literate organisations can reap a market advantage ment Each of these benefit propositions can allowed companies pursuing e-business development to drive profitability through reduced cost, greater market access, higher customer retention and more flexible production and delivery arrangements. DEVELOPMENT DRIVERS The drivers of this development across the world are many and reflect some substantive and perceptual changes in the way business has applied technology to its operations. The old days of electronic data interchange progressed with the adoption of TCP/IP meta standards and the arrival of usable and application rich web platforms. This change can be seen in the context of a wider development built upon the adoption of IT shown below. a4IT DEVELOPMENT CHART External a . pen system High bn V_— Transpareney jnjer/Intranet collaboration) g IV Ienablednew = (Network, a Business application z and communicaton) a 3 TI Value creation and 8 increased business ee 5 effectiveness = 5 < (Client Server) & II __ Increasing Productivity and & | Internal end use empowerment & | focus (Mainframe midrange) I Automation, Cost control, Data Storage Lo 1950 2020 Phase 1 saw the incorporation of IT systems to enable production automation and to facilitate a degree of cost control through large scale financial management systems. These were really an attempt to provide some economies of scale in data handling and to provide some standardization in production control. Phase 2 saw the extension of IT into enhancing productivity for the organization and IT began to impact upon end users, early office automation software, simple analysis and above all data storage capacity. The extension to new users was built upon the back of a client server platform loose linked into existing mainframe architecture. Phase 3 was the era of the intemet and the local and wide area network, where IT was used to enhance the value creation activities of the organization through customer relationship management, customer database, order and process tracking and system that facilitated JIT production and stock control/management. For the future this trend is likely to move into phase 4 with the emergence over the next five years e-businesses built around new revenue creating models that encompass both ‘bricks’ and ‘clicks’ operations. Further still phase 5 is likely to see the emergence of the truly transparent organisation built around close and distant collaboration between enterprises where the boundaries of the firm are so wide as to include a network of partially integrated suppliers, marketing agents, and customer built around common technology and shared organisational data, The phases are progressively outward orientated reflecting increasing customer responsiveness such that the final phase is a truly open and cooperative organisation with all its users and stakeholders. At the tur of the century we can see a variety of developments with some countries striving for phase 3 and 4 while others are still investing in phase 1 and 2. Japan for instance moved swiftly from phase 1 to phase 3 resulting in a high concentration of mainframe technology but with weaker process and management control systems to support their move into internet and 45intranet architecture. Each country and each region has its own IT culture and its development path but there have been some common threads that we can identify: 1. Information assets, systems, relationships ~ the building of information capability, to track customers, to invest in data systems and to value non tangible business assets. 2. Digitalisation The ability of digitial based information to transcend technology platforms, music video and voice, text easily converted, communicated and re-stored 3. Information Networks - The ability create and maintain communications across geographical and organizational boundaries, to include up stream and downstream business partners 4, Speed — The reduction in planning and strategy cycles, design to production time, customer service and delivery, product and brand innovation, 5. Intellectual Property, The recognition of intellectual property as a business asset, codified either as a patent, a process, an ability or a brand but more importantly as people with appropriate skills 6. Convergence — Technologies becoming conversant with each other and mutating across platforms driven by common standards. ‘The picture that is emerging is a very fragmented one, with high variation in countries, sectors and companies as well variation in how these key drivers have been responded to. Using this framework The e-business situation in SE Asia will be assessed followed by some key observations pointing toward the major issues for e-business development. SE ASIA AND E-BUSINESS The dominant economies in e-business terms have been the US, Europe and Japan which have produced most of the enormous growth statistics in recent years. There are however numerous reasons why growth in these economies may not represent the true potential for the rest of the world. These economies are highly developed with legacy infrastructure such as telephone, cable and a high level of IT investment and disposal income among the population. Communication infrastructure, postage and distribution systems are all in place, a highly qualified work force and economies tending toward service operation rather than manufacturing. Market Analysis The growth patterns across the world as indicated by Table 1, are confusing primarily because they reflect a development relationship. Countries in SE Asia have very different growth patterns which are masked by aggregate data. In the following (figure 2) Web Users by Region there are sharp contrasts in regional growth rates. The US is a highly developed economy and therefore growth occurs with breadth and depth, The ROW is in the underdeveloped or developing category and so growth is lower. Japan with structural and legacy issues still to resolve, is in medium growth reflecting a highly diverse economic development pattern. Asia Pacific too has areas of high growth such as Korea and low growth such as Indonesia, In summary though, The largest internet users are located in China (15.8m), 5" in world terms, followed by South Korea (14.8m), 7 in world terms and Taiwan (6.5m), 13" in world terms. Over half of SE Asian countries have only less than 3% of their population connected to the internet but in real numbers this equates to 16.6 % of world internet usage. 46Figure 2 - Web Users by Region, 1999 and 2003 Web Users by Region, 1999 and 2003 a "a us. Japan "a 33% Japan W. Europe 34% W. Europe 35% 240M Users 602M Users Source: ICCM v6.1, 2000 Analyze she Fustre In 1999 the Asia Pacific on line retail market was growing at 200% per annum and was valued at $2.8 billion (see figure 3) . This figure puts the region in third place behind the US, whose share was $36.6 billion and Europe $3.5 billion respectively. The major share of the Asia Pacific market (94% of total online retail revenue) belonged to Japan followed by Korea ($720 million) and Australia ($380 million). South Korea has seen a phenomenal rate of online brokerage whose value reached $520 in 1999, Figure 3 105 (6%) rr cee 2000According to Gartner Group B2B E-commerce in Asia is projected to reach $995.8 billion or 13.6 percent of the global $7.3 trillion B2B market in 2004, compared to $9.2 billion in 1999. This data is significant because Business to Business trading networks have less reliance upon public infrastructure, increase market activity and reach within the economy and rely upon business trust and mutual cooperation, Compare this with business to consumer where individual buyers are still connecting via home or proprietary systems, are usually located close to existing ‘bricks’ provision and are stiull distrustful of e-business activity. The East- West Siam travel company in Bangkok recently stated that “Customers are not ready to buy sophisticated travel services on-line. For the moment they will only book easy and simple services such as a hotel room or car rental.” In context, only three countries have basic e- commerce transaction facility such as secure web servers (SSL) with the majority of SE Asian consumers using email, surfing and product information. Basic infrastructure and Technology: As a summary of the current state of play for ASEAN countries, six countries have a telephone density figure of less than 5 i.e. number of phonelines as a percentage of the population. Only one country has a density figure above 40. Similarly over have ASEAN countries have PC access available to less than 5% of their population. Most ASEAN countries have less than 25% digital wireless coverage and high variability in cable access ranging from 75% to 0. With these infrastructure figures there is no surprise that B2B applications are showing the highest growth with business to consumer activity hampered by low telephone density and low connection speeds. However, on a positive note the growth in mobile networks underpins an increasing interest in WAP technology which may reduce the ing impact of traditional telephone infrastructure. Clearly, infrastructure has been a constraining factor in e-business development but wireless transmission offers an opportunity to increase diffusion in terms of the statistics, but not in terms of rich content, For this to occur, cable investment would be needed and this may well herald a growth in richness, reach and transaction value as its density increases. Skills and Human Resources Data from a recent Morgan and Banks survey indicated that SE Asia is likely to suffer a major Brain Drain in the next five years as their brightest technological talent beats a path to the US. Recent data would suggest that of the 15,000 Ph.D qualified staff recently graduated in South Korea, over 12,000 got on a plane to the US. These young workers are attracted by clear credit and recognition of their work, fast promotion, and the ability to work independently. It was interesting to note that young generation X workers in SE Asia loyal to organisations which develop their individual skills, allow them to work like entrepreneurs and challenge their ability. Salary was 5" on the list of reasons for changing employers. Staff retention will be a major organizational problem in the next five years especially in the 34-45yr age group. Looking further ahead we can only use information concerning the current situation in schools. Generally less than 30% of schools have IT access, although tertiary institutions seemingly have among the few fast and dedicated internet facilities and are currently servicing the huge demand for IT professionals. The following (figure 4) indicates very clearly that current demand for IT professionals in SE ASIA compared to countries. 48Figure 4: The percent of executive hiring for high-tech sector in Asia, Europe, Latin America and North America (1999) North America Latin America Europe ‘Source: KFI (Kom Ferry International) This situation is likely to change as government initiatives to integrate IT technology into the curriculum advances. The issue of retention though is still relevant for governments hoping to capitalize upon newly qualified technical skills and increasingly business is demanding not IT skills but broader business skills with an IT focus, able to apply IT to business and generate the economic benefits to local economies. Economic Development Issues While e-business is clearly in a dynamic state in SE Asia there are some important development issues for governments and business alike. With impressive growth statistics and buoyant revenue forecasts, companies outside Asia are looking with keen interest at the potential market place in SE Asia. This is either via direct or indirect participation. Indirect through offering goods and services via the internet in the B2C arena, and increasingly by providing the hard and soft infrastructure for B2B development. Direct involvement is through the setting up of joint venture companies but the absence of a brand focus in SE Asia is hampering the competitive potential of local businesses. Companies, with established businesses face high levels of competition with younger, but more progressive companies sporting high value overseas brands, organized online channels and relationships. In traditional business this external competitive effect has been ameliorated through joint ownership and collaboration, but where there are digital or information products there is no need for local country involvement at all. Business Development Issues Global business is undergoing a major upheaval in adjusting to a new economic environment dominated by e-business, virtuality, ubiquitous communication and customer power. However there will continue to be a need to focus activity in improving competitiveness especially international competitiveness. E-business activity can be both an ally and a threat. By making markets more transparent business activity is more widely exposed. Keeping customers loyal in this new dynamic marketplace is already stretching executives in the US and Europe. Virtual loyalty programmes such as Benz, have not made great in roads as yet but with Asian consumers used to using the internet for information search rather than purchase there are clear indications of increased rather than decreased customer mobility. When these changes in the market place are allied to IT management the issue of strategic alignment of the business becomes apparent. Markets that respond in planning cycles of months are difficult to align with IT investment strategies that are costed and implemented on the basis of years. Add to this the problems of spend justification, performance assessment and spiraling cost and business is in an unenviable position. 49SUMMARY Looking toward the future has always been an activity fraught with danger but there are some issues on the horizon for SE Asia e-business that must be considered * The low level of development of supporting infrastructure (telecommunications, finance, regulation) This is seen as a large impediment to e-business migration yet growth rates are still relatively high and wireless applications are increasing the opportunities for growth. However, even if connection is achieved there are still problems presented by local financial institutions and banking rules which make it particularly unattractive for consumers. This is evidenced by the concentration upon email, surfing and information gathering in the usage statistics. CFINet in China has the opportunity to be a powerful business to business portal but is still used only to exchange information and the actual business is still done face to face. Investment by ASEAN, Malaysia, Thailand and Korea have had a major impact upon development but * Skills shortage and labour migration I have already touched on the likely human resource problems and these will start to dominate business development. While IT skills are still in demand, a huge increase in demand for business skills such as management, marketing and business process design will need to be serviced. * Fear and uncertainty in the implementation of security and encryption technologies; Individuals and companies are worried about the encryption technology currently being applied to transactions and inter-company communications. Sensitivity to these issues is important but it is likely that security will diminish in importance as B2B operations increase in frequency and value. Recent events like the ILOVEYOU virus have heightened awareness of security again but also alerted individual consumers to the hazards of on line business. * The lack of capital to develop E-commerce: SE Asia had an enviable reputation for economic growth but the financial crisis damaged its investment attractiveness. There are signs that the up tum in growth is returning and it is likely that investment funds for e-commerce will again become available. However, governments are keen to promote this type of development and are investing heavily in Asian Portals, On line Malls and other projects to stimulate both internal and external investment in IT. * English language continues to dominate in digital environment ‘There is a natural suspicion of a dominant internet culture based upon a Euro-american English speaking model. Clearly the dominance of this culture over all others has been a result of early growth but increasingly it is likely that local languages cultures will start to emerge. The problem of course is in the practicalities of keyboard layout and usage,global versus local trading as well the danger of generating language ghettos on the intemet which do not address the global business dynamic of the 21* century. CONCLUSION South East Asia is at the threshold of a dynamic change in its business practices. The establishment of e-business as more than a business fad has caused widespread re-assessment of fundamental business principles. The SE Asian environment has the potential to engage with the world e-business community but faces a number of critical issues which are unique to this trading area. While the dominant economies have migrated to e-business with seeming ease, SE Asian growth predictions look even better but only if the focus is regional and not country specific. If Asia starts to compete within itself the major economies will be advantaged. The key to survival is to focus on key technical infrastructure, human resource development, providing a conducive and fluid economic environment and facing head on the business development issues associated with IT and marketing.APPLICATION SERVICE PROVIDERS (THE LATEST WAVE IN E-COMMERCE) Robert Kheng Leng Gay Nanyang Technological University, Singapore ABSTRACT Ever since the World-Wide-Web was launched, towards the end of the last millennium, all kinds of new businesses based on the Internet have sprouted. Some have grown at a tremendously rapid pace, while others have disappeared just as quickly as they appeared From this chaos of activity, one discernible trend is beginning to emerge: the proliferation of Application Service Providers (ASP) on the Internet. This paper will focus on this trend and will also discuss the ASP Center, set up at Nanyang Technological University (NTU) as a joint venture with the Infocomm Development Authority (IDA) of Singapore and SUN Microsystems. INTRODUCTION Internet technology has been available since the early 1970's, however, it's use was limited to a relatively small sector of the population, made up mostly of academics and researchers. Around 1988 Internet was made commercially available to businesses. However, it was not until, the World-Wide-Web became publicly available in the early 1990's that the Internet became truly ubiquitous. E-Commerce began its phenomenal growth on the back of this Some statistics demonstrating this growth are given below [US Govt. WG, 1997; Lynn Margherio, 1997} + In 1994, 3 million people used Internet. By the end of 1997, more than 100 million people were using the Internet. * As of December 1996, about 627,000 Internet domain names had been registered. By the end of 1997, the number of domain names had more than doubled to 1.5 million Traffic on the Internet has been doubling every 100 days. Cisco Systems booked over $100 million in sales on the Internet in 1996. By the end of 1997, its Internet sales were running at $3.2 billion annual rate * In 1996, Amazon.com, the first Internet bookstore, recorded sales of less than $16 million, In 1997, it sold $148 million worth of books to Internet customers, + In January 1997, Dell Computers was selling less than $1 million of computers per day on the Internet. The company reported reaching daily sales of $6 million several times during the December 1997 holiday period. * Auto-by-Tel, a web-based automotive marketplace, processed a total of 345,000 purchase requests for autos through its web site in 1996, for $1.8 billion in auto sales. As of end November 1997, the Web site was generating $500 million a month in auto sales (6 billion annualized) and processed over 100,000 purchase requests each month. While these growth figures were impressive, it was believed that that was just the beginning of more significant things to come. So much so that, on I" July 1997, President 53Clinton issued a directive to heads of executive departments and agencies of the US government sanctioning the promotion of Electronic Commerce globally [Clinton, July 1997 J. On the same day the report "A Global Framework for Global Electronic Commerce" [US Govt. WG, 1997] was released. This report included guidelines for how the government should play its roie in promoting the technology. Five types of economic activity can be identified that will fuel the growth of Internet use ‘* Expansion of the Internet : In 1994, 3 million people were using the Intemet. By 1997 100 million were using it. Experts believe that as many as | billion people will be using the net by 2005. This expansion will drive the increase in investments in computers, software, services and communications. [US Govt. WG, 1997] ‘* Electronic commerce between businesses: The use of the Internet for conducting business between businesses is expected to grow at a tremendous pace. By the year 2002, $300 billion worth of commerce, is expected to be conducted in this way. [US Govt. WG, 1997] ‘+ Digital delivery of goods and services: Software programs, newspapers and music CDs no longer need to be packaged and delivered to stores, homes or kiosks. They can be delivered electronically over the Internet. Airline tickets and securities transactions are now commonly done online. Other types of industries such as consulting services, entertainment, banking, insurance, healthcare are now becoming one of the largest growth areas. Application Service Providers fall under this category. [US Govt. WG, 1997] © Retail sale of tangible goods: The Internet is also being used for ordering physical goods Items such as provisions, books, cars and computers are being increasingly sold through the net, [US Govt. WG, 1997] + News and Advertising: The Internet is, of course, one of the most powerful media for delivering news and advertising. It crosses national borders and is independent of location and time. In recent months a lot of the glamour has gone out of investments into Internet-based enterprises. The reason is, of course, overpricing of shares of such companies and rapid and indiscriminate proliferation of start-ups in this technology. As investment markets settle to a saner level of enthusiasm, the companies, which are providing real value to customers, are being sorted out from those, which are based purely on hype. These will be the survivors who will be profitable and will grow, albeit at a less hectic pace as before. In the midst of all this ferment and disruption, new trends are beginning to emerge which will rapidly change the way businesses operate in the near future. One of these is the appearance in significant numbers of Application Service Providers (ASP) The ASP has been touted as the next Killer APP to evolve from the Internet environment, the predecessors being email and the World-Wide-Web [J. C. Bradford & Co. 2000] Estimates range from a market of US$4.5 billion [IDC, 1999] to USS20 billion [J. C. Bradford & Co, 2000] by the year 2003. APPLICATION SERVICE PROVIDERS 1. What is an ASP and what are its characteristics? The most comprehensive definition of an ASP is given by International Data Corporation (IDC). According to IDC [IDC, 1999]; 54"ASPs provide contractual service offering to deploy, host, manage and rent access to an application from a centrally managed facility. They are responsible for either directly or indirectly providing all specific activities and expertise aimed at managing a software application or set of applications," Some defining characteristics are also given by IDC [IDC, 1999] * Application Centric. ASPs provide access to, and management of applications, which are commercially available. These services are different from business process outsourcing (BPO), for instance where the outsourcing contract encompasses the management of entire business processes such as human resources or finance, They are also different from basic hosting services, where the focus of the service is management of the IT infrastructure such as networks, servers and databases with minimal applications management * "Selling" application access. ASP services offer customers access to new application environment without making up-front investments in the application licenses, server, people and other resources. The ASP either owns or has a contractual agreement with the software vendor to license access to the software, In contrast, under traditional applications management (AM) services, the customer already has acquired and deployed the application environment; the AM outsourcer takes over application management, sometimes bringing it into its own data centre. However, both AMs and ASPs do fall under the broader category IDC refers to as application outsourcing. IDC also does not consider traditional Web-hosting services to be ASP services. Under typical hosting contracts, the service provider sells (or resells) the application licence to the the customer ‘on a one-to-one basis, In the ASP model, the service provider rents access to the application on a shared basis. + Centrally managed. ASP services are managed from a central location rather than at each customer's site. Customers access applications remotely over the Internet or via leased lines, + One-to-Many service. ASP services are designed to be one-to-many offerings. The ASP partners with the vendors to package standardized offerings (providing for minimal or no customization) that many companies will subscribe to over a specific contract period Conversely, IS outsourcing, application management services, and traditional hosting are one-tone, with each solution deployed to meet the unique needs of the client organization. * Delivering the contract. The ASP is the firm that is responsible, in the customer's eyes, for delivering on the customer contract, ensuring that the application service is provided as promised. ASP services will often involve several partners. If a problem arises, however, it is the ASP that is responsible for closing the loop on the problem ticket, even if the ASP ‘works with other companies to provide the actual support. 2. Types of ASPs There will be as many types of ASPs as there are types of applications, However, IDC {IDC, 1999] offers the model shown in Figure 1 With reference to figure 1. The definitions of the different levels of complexity (Y-Axis) are, * Analytic applications. These applications include any application built to analyze a business problem (e.g., financial analysis, customer churn analysis, Web site analysis, and risk analysis)* Vertical applications. These applications include any industry-specific application, such as MRP in the manufacturing industry, patent billing in the healthcare industry, and claims processing in the insurance industry. + ERM applications. These applications include accounting, human resources, materials ‘management, and facilities management. + CRM applications. These applications include segments such as sales force aytomation, customer service, and marketing applications. * Collaborative applications. These applications include groupware, email, and conferencing applications. + Personal applications. These applications include office suites such as Microsoft Office and consumer applications (e.g,, games, home productivity, and "edutainment"), Figure 1 The ASP Market Landscape “Lower volume business ‘More customization * Greater business eomplesity + Need to integrate with ‘other bosiness systems smplex sales cyle Extent of ASP Services Provided Soute Iteration! Data Corparaton, 1988 The definitions of the components in the; extent of ASP (X-Axis) are: * Core services. These services include the foundation services that an ASP needs to provide in order to manage the application environment and provide a base level of customer satisfaction. These services include services such as application updates and upgrades, 7x24x365 monitoring of the application, network and servers on which the application would run, and basic customer support. © Managed services. These services include all of the core services, plus additional services and guarantees around support, security, application performance, and data redundancy They would include services such as SLAs around application performance and data security, dedicated technical support personnel, and daily backup of the application and its data‘© Extended services. These services include all the managed services, plus additional services. Although the extended ASP services begin to border on a custom-delivery model. Extended services include services such as application configuration and extension, strategy and planning and educational support. 3. Competencies required To become a high quality ASP requires a wide range of expertise and skill sets. These can be classified under four categories; Business, Computer Systems, Data Management and Application Domain Expertise. 3.1, Business Business model Project management Services infrastructure Business development Customer support 3.2 Computer Systems (Hardware) Clients and Servers Networks; wired and/or wireless Data storage systems Security Customer support Quality of Service monitoring 3.3 Data Management (Software) * Database management * Web-site design/maintenance * Electronic security * Customer support 3.4 Applications License administration Application integration Application management Application support Application sales expertise This is a wide range of skills and it is unrealistic for an ASP to cover all of these alone. An ASP has to determine its position in the Market and select the right partner: to run its business. 57the market. 4. The positioning of ASP: Because of the wide range of competencies required of an ASP. It is unlikely that an ASP would do everything itself. Focussing on delivering the applications, most ASPs would go into strategic partnerships with vendors and other service providers to provide their particular services, Figure 2 (IDC, 1999] gives a comprehensive picture of where an ASP should be positioned. ‘Distributors Resellers Service Firms ltwork Providers + sis + ISPs + Consultants / + Telcos / 5 Outsoureers * Hosting Application Vendors. Source International Data Corporation, 1999 Figure 2, ASP Positio WHY ASPs? The ASP has been touted as the next Killer App to evolve on the Internet . But, why has it become such an attraction for investors and what value do they bring to their customers? Apart from the technology push, three strong reasons are driving this trend, they are necessity, attractiveness as a business and strong value propositions, 1. The necessity The emergence of the ASP is really a consequence of the difficulties faced by smaller companies of exploiting the advantages of implementing IT. Three of the most cited disincentives are [J. C. Bradford & Co., 2000] ‘+ The high up-front cost of a packaged software license. * The lack of inhouse IT human resources to install and maintain the software 58© The high cost of building and maintaining the IT infrastructure necessary for the application The ASP of conducting business overcomes these obstacles for the smaller companies whilst making it cheaper and more efficient for the larger companies. 2, The attraction for the investor From the point of view of the investor, the following are compelling attractions A high percentage of recurring revenue Leverage generated from a one-to-many solution model. High switching costs for customers. High returns on investments. Ability to sell high-margin, value added services into the customer base. 3. The Value Proposition of the ASP. Turnkey solutions, Predictable costs and lower up-front investment Faster time to market. Ability to scale rapidly. Easier upgrade cycles. Minimizing of IT human resource requirement. A viable solution to a mobile, distributed workforce. Improved focus on core business competence. While the promise of better service and lower costs, through outsourcing to an ASP is real, customers need to be aware of what they require in the Quality of Service they engage the ASP for. Hence the need for careful drafting of Service Level Agreements (SLA), SERVICE LEVEL AGREEMENTS In the age of Internet, customers will demand 100% in all aspects of service levels. The most important being; reliability, security, availability, accessibility, and rapid response. In developing Service Level Agreements with Customers, ASPs need to be open and honest so as to be able to deliver when required. It is impossible to deliver 100% for every aspect of service offered, but a high percentile (usually in the 99% + range) is mandatory to attract any business at all. Different customers will have different priorities in the level of service provision. To some a high degree of security is most important while to others a fast, response is more important. ASPs need to tailor each SLA according to the customers’ requirements. Most ASPs will rely on several alliances and partners to run their business. So, there is no avoiding dependence on others in the quality of service to be rendered. This basically means they should select their partners carefully. The key phrase here is "best of breed”. In the end, it is the ASP that is responsible to its customers, not the partners. ‘THE ASP CENTRE AT NTU Although there are now companies in Singapore offering content hosting and service co- location services, there is still a huge gap in scope and depth between them and those found inthe US, such as AboveNet and Exodus. Intending Singapore to become a major content and services hub, the Government has already taken steps to build an extensive physical infrastructure. With this infrastructure in place and operational, it is now time to start nurturing world class service and application providers. Recognizing this, NTU, IDA and SUN Microsystems have initiated a joint venture to form the Application Service Providers Centre (ASPCentre), This was launched through the signing of an MOU by NTU, IDA and SUN Microsystems on 27th January 2000 The objective of the centre is to provide a central point where aspiring ASPs, can obtain assistance in the form of R&D (NTU), infrastructure expertise (NTU and partners), business advice (NTU and partners) and venture capital (VCs and possibly NTU). This is the first centre of its kind in Singapore. The ASPCentre will also be available to infrastructure providers and service providers in Singapore for piloting new and innovative services. The ASPCentre will form strategic alliances with the best of these ACTIVITIES OF THE ASP CENTRE Activities can be simply divided into two major categories, R&D and ASP development, including; incubation and deployment. The activities and of the centre will be of a consultative nature, based on project agreements. There will be 3 senior Service Provider (SP) Architects leading a team of six other project officers. There will also be six research engineers developing new technology under the supervision of NTU staff. 5 Scholarships for graduate students will be offered under the programme. The organization of the Centre is shown in figure 3. The strategy of research, development and deployment is illustrated in figure 4. Management Board Director Operations Manager Academic Committee [ 1 Business Industry anes Associates Development Development Penrice Staff and Manager Manager PG students Systems 3 SPA Education Administrator] Consultants MBA/MSc Courses /Projects FYP Projects Technician | Support/Back-up ! Figure 3, ASP Organization Chart 60Types of ‘Types of Core Activity Manpower [Development | Development | Consultancy ] Execution of | Market and of New and | of New ASP Business Improved Applications Industry Development Technology Projects University | High High Medium Low Medium Staff Postgrads, | High High Low Medium Low REs and Scholars SPA Low Medium High High Medium Consultants Project Low Medium Medium High Medium Officers ‘OM, BDM__| Low Low Medium Medium High and IDM Time Tto3 years [3monthsto | Twotosix | 2tomonths | Immediate to Horizons 1 year weeks 2 years Figure 4, Manpower A Focus Chart KEY FOCUS AREAS 1, Research and development As indicated in figure 4, most of the R&D in new technologies and applications will be done by University staff and graduate students, Although the leadership of the centre will be provided by the School of EEE, the centre is a multidisciplinary one and will involve several Schools at NTU, In particular, the Nanyang Business School and School of Computer Engineering are expected to play major roles in the activities of the centre. It will be a focal point for R&D and consulting activities involving the Internet at NTU. ‘The areas of current interest are: Business Modelling of ASPs. Architecture for ASPs Security technology and systems Intelligent agents for data mining, knowledge discovery and environmental scanning Web-based collaborative workflow and project management Web traffic management and transaction control Real-time and embedded systems Wireless applications 2. ASP development While the marketing and management side of the Centre will drive the business aspects of this, the technological part will be led by the Service Provider Architects. The roles of the SP Architects are similar to those of their counterparts in the Building Profession, They will, in consultation with clients, design internet-based systems which will provide the business services they intend to offer. The actual building of the systems will be done by software 61engineers and programmers. The main difference is that prototypes of the services can be built and tested at the ASPCentre before being launched live on the Web Initially the Centre will focus on the following activities to address the needs of industry. * Joint development with local industry in new and innovative application services * Provide training and support so as to equip local industry with the capability of delivering quality services on the internet. * Provide branding and certification of service providers, The branding will signify qui and reliability. The SunTone program will be used as a basis for certification, * Pilot new generations of Internet Protocol network services Eventually, the center could provide the best practices to help the aspiring service providers achieve high level of reliability and availability. The centre will have sufficient computing capacity to host 30 SP projects at any one time However, it would be impractical to provide space to house 30 projects. The intention is to house about six projects at the centre at any one time. Other projects can be initiated at the centre and then moved offsite CONCLUSION The Internet has spawned a whole new breed of Application Service Providers which will change the way business everywhere in the world is conducted, significantly, in the near future. It is timely that Singapore is now gearing up to address, this challenge. The joint venture between NTU, IDA and SUN is indicative of the degree of seriousness in which the University, Government and Industry view this in Singapore. This is one area in which both Governments and Universities can piay a significant role in helping and promoting industry REFERENCES Clinton, W. J., "Memorandum for the Heads of Executive Departments and Agencies", Presidential Directive, The White House, Office of the Press Secretary, July 1997. IDC Corporation Bulletin, "The ASPs' Impact on the IT Industry: An IDC-Wide Opinion", Document No: 20323, September 1999. J. C. Bradford & Co., "ASPs : The Net's next Killer App", Information Technology Solutions Report, Extending the Enterprise, 2000. http:/Awww.aspisland,com/trends/jcbradford. Lynn Margherio et al, "The Emerging Digital Economy", US Department of Commerce Report, 1997. http://www.ecommerce.gov/aboutthe. htm, US Government Working Group Electronic Commerce, "Framework for Global Electronic Commerce", First Annual Report. July 1997. http:/Avww.ecommerce.gov/frmewrk. htm, 6230 YEARS OF SOFTWARE ENGINEERING : FROM ART TO INDUSTRY ? ‘Anne-Marie Hugues ESSI - Université Nice Sophia Antipolis 930 route des colles - BP 145 - 06903 Sophia Antipolis Cedex France 1. KEYWORDS: Re-use, Quality factors; Workflow, Process management; Knowledge management ; Software Components, Software Life cycle, Traceability , Interoperability, SPICE, CMM, ISO 9000 2, ABSTRACT: The software production process has been tentatively improved during 30 years in both directions: reducing costs and delays and increasing quality and productivity. Today, the advances of technology allows us to build software from "ready to use" components. Are the organizations prepared for this new challenge? We focus on programming from software components, according to quality standards (1809000, SEI CMM, SPICE project...) and discuss the opportunity of their strict application in this context. We describe the necessary reorganization of the software production process while creating or reusing components. ‘We emphasize the necessity of defining a specific workflow for software production as it is done in any other industry. We conclude with an introduction to software knowledge management 3. INTRODUCTION: STATE OF THE ART AND TALK POSITIONING The term Software Engineering has been invented by NATO in 1969, it could be perceived as a kind of provocation at a time when Donald Knuth was publishing his "Art of computer programming"[Knu 68]. Is the software factory which was launched 15 years ago [Eureka] actually existing today? Why has the "Mythical man- month” [Brooks 75] been republished 20 years later with slight amendments? We believe that the software crisis of the 70s is not quite over. ‘As we find new software technologies we must invent the organizations that will support them and not merely reproduce organizations adopted by other industries, We are concerned by two classes of organizations: - Companies that have to solve an IT problem; they may either have a specific software application built by their own computer department or by a software house; they may also buy (or rent) and customize an appropriate package , i. ¢. E.R.P. = Software houses whose business is producing and selling (or renting ) software products. 63We focus on the second class. These software houses whose business is building software are confronted as any other enterprises to deliver a good quality product with a high ratio of productivity. As time goes by, these companies have learned how to re-use code, then classes and now components, but we argue that re-use is not only subordinated to technical knowledge but quite as much to a managerial problem of creating an organization that will facilitate continuous improvement, knowledge management , quality and productivity. 3.1 Mastering the process ‘Mastering the software development process’ will be the most dramatic stake as far as quality and productivity are concerned. It means meeting deadlines, insuring control responsibilities, and checking the conformity of deliverables while avoiding cost overruns. The software process starts with the elaboration of requirements (input), ends with the production of the software product (output), or more precisely with the disposal of the software product if we take into account the maintenance phase of the software product; the resources are both human beings and tools (hardware and software -including software components- ), the constraints are multiple : project planning, (delays, costs) quality factors , environment (organizational, economical, political, technological -including the legacy system- ....) The software process can be decomposed in two ways: step wise refinements or multiple iterations, i.e. Cartesian approach or empirical approach. ‘The first step made towards organization in software engineering was the classic waterfall model [Royce 1970]. The idea was to copy what was done in manufacturing industries applying step wise refinement : plan what you intend to do, do what you have planned, and then make sure you have done right at each step of the process... Different phases of the process have first been pointed out, each defining a level of abstraction:, requirements specification, implementation, validation... The V life cycle and other derived models have enriched this approach without radically changing it.. This way of working derives directly from the abstract way of thinking of engineers working in manufacturing industries (chemistry, mechanics...) After some 10 years practice of this classical life cycle models, it has been taken into account that software is not a manufactured product and as so software impacts its environment and evolves continuously according to that environment, software must be designed in strict accordance to users requirements who most of the time change their mind.... The iterative or empirical class of models has been proposed to undertake these two elements. The Spiral model [Bochm-88,] describes development as an iterative four phases process where various domains are taken into account : expression of needs, feasibility, prototyping and development of the final product. Nearly at the same time (carly 90s) a new term "Rapid Application Development" (or RAD) appeared whose origins sprang from the frustrations of users and the fast moving IT business. RAD grew as a movement in a very unstructured way. Many tools and environments were sold invoking the RAD process which was not commonly agreed. Lots of users who had confidence in that approach have been disappointed when maintenance time arrived on products quickly designed but poorly structured. Since 1993 the DSDM [DSDM]J consortium has tried to structure the RAD landscape by standardizing the RAD method . 64At the same period, as object oriented programming was growing, it became obvious that the different phases previously exhibited in a classical life cycle had to be enriched in a pure object-oriented approach to software development, the “fountain” model " [Henderson 93] incorporates reuse and domain analysis/design. More recently Rational Unified Process [Rational] is candidate for becoming a standard. 3.2 Qui ity standards applied to component programming Quality is managed by two types of models : certification and maturity models. Certification models They intend to detect the non quality and to correct unconformed products before they are sold. Ex: ISO 9000 [ISO 9001], DOD 2167A, IEEE [IEEE 1987]. The certification is a binary decision. Maturity models They are used to measure the ability of a provider to develop and sell a software product; Let us quote the Capability Maturity Model [Humphrey 1989] recommendations and the SPICE project [SPICE] ... Quality requirements - Necessity of communication among developers ~ Necessity a of communication between users and developers ~ Necessity of control and visibility during the whole life cycle ~ Necessity of rigorous and frozen requirements before design and implementation ~ Efficiency and accuracy ~ Reliability - Testability = Maintainability - Flexibility - Traceability ~ Respect of costs and calendar ~ Commitment and training of people CONCLUSION When changing for a component based programming we shall take into account these quality standards and adapt them at the new context. We shall mainly find an organization and a process management that fits the new way of production, 4. CHANGING FOR A REUSE POLICY WHILE MAINTAINING (OR INSTAURING ) QUALITY SYSTEM As in any other industries, we want to reuse the work done before in order to decrease costs and increase quality We are progressively learning reuse. Reuse can be done at different levels : informal reuse of code, reuse of objects and classes and libraries, mature reuse of components that have been designed on purpose ; but we can also reuse documents: design, architectures [Gamma95}, graphical user interface... 65Component life cycle Figure 1 : component life cycle In practice, reuse can be opportunistic or planned, it has much more chances to work if it has been planned in advance i. if pieces to be reused have been designed on purpose; There are two ways of approaching reuse: bottom up ie build pieces that you will reuse eventually, or top down build the assets that you need in a project according a predefined policy. We think you will have more chances to reuse them in that context. The kind of reuse we address regards reuse of software components over various products of the same family, in other words over a product line; We don't think we are ready to reuse off the shelf business components over various products and various organizations as we think we are not yet mature to actually maintain the off the shelf business components. Classes, objects , components Before discussing the reuse mode, we have a closer look of what a component is. It depends if we look at theory or reality. As defined by researchers, a component should be a black box, should have a narrow interface and a single point of access, few variation points rather configured during instantiation. But if we look at what industry is actually working with, we see that an object oriented framework is called a component i.e. a large piece of code with some structure ; interface is provided by entities i.e. classes in the asset; the variation is implemented through configuration or specialization or replacement of entities in the asset. The theory says that we can buy component on markets, but industry prefers to develop them internally as experience shows that the one you buy have to be considerably adapted to fit ‘one's need. .. Researchers tend to focus on functionality but practitioners take into account all quality attributes (performance, reliability, maintainability..). 66Software Life cycle Component oriented process Quality ‘standards| Quality smanval Fig 2 component oriented process So, as far as industry is concerned, a component should be developed with the same quality criteria as a product would be. We can apply to component development the same quality standards as we apply to quality products. A component life cycle is described in Fig 1 This component life cycle is integrated in a larger one that is the product life cycle , itself included in the product line life cycle. So when looking at the whole software life cycle we get a process as described in Fig 2. Al this process has to be supported by organization. We suggest that the one adopted should be a matrix organization that = reflects a team-based culture, - allows continuous improvement - and facilitate knowledge management. Each project should get team members from a knowledge domain , one of this knowledge domain should be "reuse". Matrix organization jram management aes) > | vane = an —-, car fee T ae _ ~ Organization roduc Fig3 Matrix organization 67There should be a person in charge of reuse whose job is to maintain the components dedicated to reuse. As for software, organization can be thought of in terms of patterns [Ambler 98 99] organizational patterns}. 5. CONCLUSION : COMPONENT BASED SOFTWARE ARCHITECTURE The first thing to do when implementing reuse is to define a software architecture in which the components will be developed and later fit into. A software architecture can be viewed as a structure which comprises - Design (maybe using design patterns that are design components) ~ Quality attributes (functional and non functional) - Components and connectors (code black boxes) Working spaces. A software product is often part of a product line. For instance in a software house working on corporate software, requirements generally overlap. It is suitable that all these products should be developed according to the same architecture and using the same components. The difficult partis to find the adequate level of variability versus commonality ‘We recommend to only implement the expected variability and balance carefully investment and risk. As a conclusion we can state some thumb rules when implementing reuse ~ foresee in the long term ~ adapt architecture incrementally ~ work on actual fact ~ take into account the cultural aspect, ~ train people - measure progress to encourage In order to fulfil these rules, itis suitable to use tools methods and organizati the re-engineering of the process. We propose [InterFas] from FAS NET whic a connected tutorial mn that support described in 6. BIBLIOGRAPHY [Ambler 98] Scott W. Ambler , Process Pattems, Building Large-Scale Systems Using Object Technology, Cambridge University Press/SIGS Books, July 1998 [Ambler 99] Scott W. Ambler , More Process Patterns Delivering Large-Scale Systems Using Object Technology, Cambridge University Press/SIGS Books, July 1999 [Blum92] B.I. Blum Software Engineering , A Holistic View Oxford University Press, 1992 [Boehm 81]B.W Boehm, Software Engineering Economics, Prentice-Hall, 1981 68[Boehm 88]B. Boehm, A spiral model development and enhancement IEEE computer21, may 1988 [Booch 94]G Booch 1994, Object oriented analysis and design with applications, Benjamin Cumings 1994 [Brooks 95] F Brooks The mythical man month Addison Wesley (second edition republished 1995) [CVS] : http://www cyclic.comveyclic-pages/evsdev.htm! [DSDM} : http://dsdmtest software.ple.uk/ [Eureka } http://www eureka be/Home/ (Henderson 93]B Henderson-Sellers, JM Edwards, The object oriented systems life cycle, Comm. of the ACM 33, September 1990 (chap 3 and 9) [Humphrey 89]W S Humphrey , Managing the software process, Addison wesley 1989 (chap 3) [Harmon 98]Paul Harmon and Mark Watson Understanding UML, the developers guide 1998 [IEEE 1987] IEEE 1987, Standards for software project management plans, IEEE 1987, chap 10 [InterFas] http://www interfas.net [ISO 9001,1987}, ISO 9001, Quality systems - Models For Quality Assurance In Design Development Production Installation Servicing, International Organization For Standardization Geneva,1991 {ISO 9000-3,] ISO 9000-3Guidelines for the Application of ISO 9001 to the Development Supply And Maintenance of Software , Intemational Organization For Standardization, Geneva 1987 [ISO 1995] 1207-2 Information Technology : Software life cycle process part 2 Configuration management for software {Gamma 95] Erich Gamma, et al , Design Patterns : Elements of Reusable Object-Oriented Software Addison-Wesley Professional Computing October 1995 {Knuth 68 JD Knuth The Art of Computer Programming : Volumes 1,2,3, Third Edition ‘Addison Wesley Longman [Meyer 2000]B Meyer, Object-Oriented Software Construction, 2nd Edititon, 2000, Prentice Hall [Metzger 96] PW Metzger, John Bodie Managing a programming project, Prentice hall 1996 [organizational patterns] http://www.bell-labs.com/egi-user/OrgPatterns [Oskarson 96}Osten Oskarson, Robert L glass, An ISO 9000 approach to building quality software, Prentice hall 1996 [Pressman 87]R. Pressman Software Engineering: A Practitioner's Approach Mcgraw-Hill, 1987 [Rational] http://www rational.com {Royce 89]WW Royce, Managing The Development Of Large Software Systems Concepts ‘And Techniques, Proceedings of the 11 International Conference on Software Engineering Pittsburg May, 1989 69[Rumbaugh 91] Rumbaugh, Blaha, Premerlani, Object Oriented Modelling and Design , Prentice Hall 1991 [Shlaer 88] Shlaer , Mellor, Object Oriented System Analysis: Modelling the world in data, ‘Yourdon Press 1988 [Schach 96]Stephen Schach, Classical And Object Oriented Software Engineering, Irwin 1996 [Sommerville 96] Ian Sommerville, Software Engineering And Its Application, Prentice Hall 1996 [SPICE] http://www.sqi.gu.edu.aw/spice [RCS] WF Tichy, RCS a system for version control, Software Practice And Experience 15, july 1985 [Yourdon 79] E. Yourdon, L. Constantine, Structured Design, Prentice-Hall, 1979 [Workflow]: http:/Awww.aiim.org/wfme/mainframe.htm 70Informatics: A Truly Interdisciplinary Science — Prospects for an Emerging World — Dines Bjgrner Department of Computer Science & Technology Technical University of Denmark DK-2800 Lyngby, Denmark E-Mail: db@it.dtu.dk July 1, 2000 Abstract Informatics is a science and information technology implements informatics appli- cations. Mastering informatics thus is a crucial, but intellectually taxing prerequisite for meaningful information technology. ‘Todays information technology uses suffers from a serious lack of understanding. the difference between informatics and information technology. In this article we will define the notions of informatics. We will emphasise the engineering of domain descriptions, informally and formally — as prerequisites for software requirements and software design engineering. We close by speculating on the consequences of what this article is about for academic, IT consuming, and IT delivering, commercial institutions in the Far Bast. 1 Introduction 1.1 Summary We define Informatics as some conceptual “sum” of mathematics, computing science, and applications — as implemented by information technology. We define engineering as ‘walking the bridge” between science and technology. We then bring (seven) cursory, non-technical examples of applied informatics from the infrastructure components of (1-2) transportation (railways, roads, air transport, shipping, container logistics), (3) the health- care sector (from rural nurse via village clinics, physicians, hospitals, pharmacies, medical insurance to government regulatory agencies), (4-5) financial service industry (banks (of all 7kinds], insurance companies, brokers, stock and bond exchanges, portfolio mgt., clearing, regulatory agencies, etc.), (6) fisheries industry (from the fishing grounds via fish auctions, processing, transport and retailers), and (7) the “consumer” market including suppliers, traders and producers. In this brief survey we emphasise the interplay between mathemat- ical modelling, software development, relevant groups of stake-holders (procurers {usually enterprise strategic management], tactical and operational management, workers, clients, TT and other resource providers, “the public at large” , regulatory agencies, and politicians), and technology. The aims & objectives of the talk is to advocate that institutions like the AIT (The Asian Institute of Technology), UNU (the UN University), regional universities, and other institutions in developing countries and in countries in transition, that such in- stitutions, when planning so-called IT-strategies, (re-)focus on informatics, a stable view, and that such a focus can be followed up by less technology and more concept drive, ie. shifted away from short term, unstable concerns to longer range, stable concerns. The article is “carried” by a number of formal models — which are brought only to indicate what we mean by such formalisations, not in order that they be immediately understood. 1.2 Background Informatics is confused with Information Technology. IT is usually taken to stand for Information Technology. But Informatics is not just technology — not “gadgets” alone. Informatics is a subject field, is a discipline, and has a rich variety of complementary theories. etc. Information Technology is “hard”: Builds on the natural sciences, notably electronics and materials sciences: Electronic chips, fiber optic cables, etc. Informatics is “soft”: Builds on mathematics. In this paper we wish to analyse the above and suggest an alternative view on Information ‘Technology: As the technology that implements Infor- matics, just as bio-technology implements parts of our need for medical cure. Information ‘Technology represents a universe of material quantity: Faster, bigger, smaller, cheaper, etc. MHz, giga bytes, pocket-size, 99 USS, ete. Informatics represents a universe of intellectual quality: Better, fit-for-purpose, pleasing, adaptive, usable, correct, etc. 1.3 Informatics: Sciences & Engineering Informatics is here seen as some “sum” of Computer & Computing Science, Mathematics, and Applications. Computer science is the study and knowledge of the phenomena that can exist inside computers: Data and processes. Computing science is the study and knowledge of how to construct those devices: Data and processes. Software engineering is, to us, the triptych of engineering understandings of the domain of applications — void of any reference to computing (&c.), of engineering requirements to software for computing applications in the support of operations of the domain, and of engineering the design of that software. An engineer “walks” the bridge between science and technology: Designs technology based on scientific insight; and analyses technology to ascertain its 72science content. 1.4 Infrastructures & their Components By a country’s or a region’s infrastructure we understand the mechanisms that enable societal operations, ie. infrastreuture is a socio-economic concept. By an infrastructure component we understand an aggregation of enterprises which facilitate “these” societal operations. Examples of infrastructure components are: The healthcare sector, the trans- port sector (road, rail, air, shipping), the financial service industry sector (banks, in- surance companies, brokers, securities & commodities exchanges, etc.), education, the consumer/supplier/trader/producer sector, &c. Operations in these sectors increasingly use computing & communication. IT within all infrastructure components is fragmented: Hundreds of software pack- ages “serve” most of these infrastructure components. But these software packages rarely “communicate” with one-another: Cannot share data, cannot mutually invoke functions, and often require different computing & communication platforms. ‘The software pack- ages rarely reflect a common understanding of the relevant, “underlying” infrastructure component. That is: To properly exploit computing for very large scale, socially ben- eficial infrastructure applications, we must [“first”] research & develop models of these infrastructure components. 1.5 Informatics Development In developing computing, and hence communication, support for any application, we con- struct and analyse (develop and research): descriptions of the domain, descriptions of the requirements to desired software, and descriptions of the design of that software. That is, development results in the following refinement documents: Domain descrip- tions which indicatively specify the world as it is, requirements descriptions which putatively specify the world as we would like to see it, and software design descriptions which “imperatively” specify the software as it will be. All descriptions are developed in phases, stages and steps. The last refinement step “is executable” ! 1.6 Informatics Development — Some Observations The domain descriptions need be both informal and formal, ie. mathematical. The for- mal descriptions make use of a variety of mathematical and formal description ‘methods’. There are now an abundance of ‘proven’, mathematically sound description and refine- ment techniques. Some description techniques focus on information and its functionalities: VDM-SL, Z, B, etc., others emphasise temporality: ITL, STeP, the Duration Calculi, etc., yet others emphasise states, concurrency and behaviours: Petri Nets, StateCharts, CSP, es, etc., and yet others emphasise compositionality, separate documentation, etc., Some techniques attempt to embody two or more of the above: RAISE, etc. Descriptions require 73analysis, validation and verification — and techniques like logical theorem proving and model checking illustrate the mathematical nature of informatics development engineer- ing. Informaties, as other engineering disciplines, thus makes good use of mathematical principles. 2 Examples of Infrastructure Informatics 2.1 Models We illustrate some infrastructure components. We do so by hinting at their descriptions: Through informal text; through “pictures”: A picture is worth a thousand words; and through formal text: A formula is worth a thousand pictures. The purpose of showing the (seven) examples is: To indicate what we mean by infrastructure component. To indicate what we mean by their description: informal and formal. To show what it takes to describe and formalise. And to indicate that it can, indeed be done. The seven examples are: ‘Transportation: (i) Air and (ii) rail, (iii) healthcare, financial services: (iv) bank and (v) stock exchange, (vi) fisheries industry, and (vii) the “consumer” market All examples will, naturally, be “grossly” simplified ! 2.2 Transportation ‘We exemplify two kinds of traffic. Air traffic and train traffic. There are some similarities, and there are some differences. The models are “rough sketches”: We model only the traffic net, timetables and traffic — and hints at scheduling functions. We can, easily, extend these models to cover “all” additional transport notions: Passenger and freight handling, transport vehicle despatch, monitoring and vehicle and net control. &c. 2.2.1 Air Traffic Airports, a:A, airdomes, d:D, air corridors, w:W, flights, EF, and aircraft positions, p:P are further undefined entities. So are times, t;T, and airport, an:An, and flight fn:Fn, names. An airspace, as:AS, associates with every airport an airdome and a set of non- stop destination airports and their air corridors. A time-table records, for every flight, the arrival and departure times at origin, stopover and destination airports. Air traffic is a function from time to airspace and positions of aircrafts. Scheduling assigns to a time-table and an airspace a possible infinite set of air traffies — said to satisfy the time-table. type A, D, W, F, An, Fn, T, P AS=A wp (Dx (An q W)) TT =Fn w (An w (T xT) 74TF =T (AS x (F w P)) value scheduling: TT x AS -> TF-infset ‘The above expresses an essence of air traffic. It is based on such essential models that, comprehensive theories of the man-made universe of air traffic can be researched and applied, 2.2.2 Railways A railway net, n:N, consists of lines, LL, and stations, s:S. Lines and stations consists of rail units, u:U, and rail units have 2, 3, 4 or more connectors, e:C Figure 2: A Rail Net Stations have names, sn:Sn. A path, p:P, is a pair of connectors. A state, o : E, of a rail unit is a set of paths. 75type N,L, $, Sn, U, C, TN value obs.Ls: N+ L-set, obs.Ss: N + S-set obs Sp: L + Sn x Sn, obs Us: NIL|S + U-set obsCs: U + C-set type P=CxC, D=P-set, = D-set R=(UxP) TT =Tn w (Sn w (T xT) TF =T + (Nx (TN w R)) value scheduling: TT x N > TF-infset A rail unit state space is the set, w : ©, of all the states it may attain. A route is a sequence of rail units and an [open] path. A time-table records for each train number the arrival and departure time at stations. Traffic is a function from time to rail nets and positions of trains. Scheduling assigns to a time-table and a rail net a possible infinite sets of traffics — said to satisfy the time-table. The above expresses an essence of railways. It is based on such essential models that comprehensive theories of the man-made universe of railways can be researched and applied. 2.3 Healthcare A healthcare sector consists of many stake-holders: Citizens, medical doctors, nurses, hospitals, clinics, pharmacies, health insurance companies, the pharmaceutical industry, and government. value Healthcare(cs,ms,ds,js,ns,hs,ks,ps,g) = (es) || Moms) || D(a) | 1G) || Nas) IL H(bs) |] Ks) || P(ps) || GC) I (eit) | es }, Mims) I[{ dru(a) | dds }, 1s || { nur(n) | nns }, H(hs) I {elie [ks }, "P(ps) { med(m) | m:ms }, II { ins() | isis }, { hos(h) | hchs }, Il { pha(p) | p:ps } channel { em[x,y}:J | x:Cx, yMx }, { ed[x,y]:J | x:Cx, y:Dx }, {ie[x,y]:J | x:Cx, y:Ix }, { en[x,y]:J | x:Cx, y:Nx }, etc. 76Figure 3: A Healthcare Sector Network 6 healthcare process graph {dglx}:R | x:Dx }, { mg[x]:R | x:Mx }, etc. The above expresses an essence of healthcare. It is based on such essential models that comprehensive theories of the man-made universe of healthcare can be researched and applied. 2.4 Financial Services 2.4.1 A Bank Banks, b:B, have clients, k:K, and clients have (and may share) accounts, a:A, through account numbers ¢:C. The bank records sharing of client accounts, status of these accounts, and processes transactions “out of a” repository, r:R, into which all client and all (“own”) bank transactions are inserted. ‘Transactions can be posited for processing in any time interval: Now or in future. The bank “cycles” between serving clients and processing transations. type K,C,A B= ({clients} m (K mp C-set)) U ({accounts} m (C m A)) U ({epository} wR) R=(T x T) w@ Jobs, Jobs =C 7 Trans-set ‘Trans == mk-Trans(rn:Rn,vi:VAL*) value cycle: B 5 Unit cycle(8) = let f" = client step(8) [] bank step(8) in cycle(6") end i7value client step: B B client step(8) = let (c,tr) = client.ch?, t = clock.ch? in client(c,tr)(t)(8) bank step: BB if B(cepository) = [] then # else let t = clock.ch? in if is ready Task(A)(t) then let (((t’t"),¢,mk.Task(m,al)),6") = sel_rmv-Task()(t) in let rout:Routine + rout € (f'(conditions))(rn) in let (9’x) = E(c,rout)(al)(t,t't")(6") in insert(r)(6") end end end else 6 end end end end The above expresses an essence of banks. It is based on such essential models that com- prehensive theories of the man-made universe of banks can be researched and applied. 2.4.2 A Stock Exhange Stocks, named 6:8, are offered, 0:0, bought and sold. Offered stocks may be withdrawn. Sold stocks are cleared. Offers, ofrs:Ofrs, state time period of offer, quantity, low and high accepted or bid prices, etc. A stock exchange reflects the above, and the market is a function from time to the state of the stock exchange. Figure 4: A “Snapshot” Stock Exchange View of Current Offers of a Single Stock 78type SE, Buy, Sil, ClRm, S, O, Ofrs, Ofr, T, Q, P, R, Clrd, Rmvd SE = (Buy x Sil) x CIRm Buy, Sll=S yp Ofrs Ofts = O yw Ofr ! Oft = (TXT) we (Q x (lo:Pxhi:P) x ...) CIRm =O 7 Clrd | Rmvd Clrd = S x P x T x Offs x Ofrs Rmvd = $ x T x O x Ofr Market = T > SE value place: {buy|sell} xBxQxSx (o:P xhi:P) x (bt:Txet:T) x... + SESE xR wthdrw: O x T+ SE4SExR next: T x SE + SE next(t,se) = if pelr(t,se) then felr(t,se) else if prmv(t,se) then frmv(t,se) else se end end pelr: T x SE > Bool, felr: T xSE > SE prm: T x SE ~+ Bool, frm: T x SE + SE The above expresses an essence of stock exchanges. It is based on such essential models that comprehensive theories of the man-made universe of stock exchanges can be researched and applied. 2.5 Fisheries Industry We model just a fragment: Fishing boats arriving and departing from a fish auction quay. The fish auction facilities: trucks moving incoming, to-be-sorted, sorted, auctioned and sold fish boxes; the fish box storage areas: incoming, sorting, auction and sold. the sorting; the auction; and the auction management: sellers and buyers, etc. We indicate interfaces to regulatory agencies. &. type Fldx, Cldx, Idx, Bldx, SAldx, Tidx Boats = FIdx yw Boat Fidx yp Cldx Coops = Cldx 7} Coop 19Figure 5: A Fisheries Auction Harbour &c. eZ 2== ~*« Buyrs = Bldx q Buyer value boats:Boats, coops:Coops, buyrs:Buyrs, ficop:FiCop value system: Unit + Unit system() = '* agents «/ || { isher(fi) (cop(f))(boats(A)) | eFax } | || {cooperative(e)(coops(c)) | e:Cldx } | || { inspector(i) | i:TIdx } |] auction master() |||] { buyer(b)(buyrs(b)) | biBldx } /+ components +/ || in_buffer() |||] { sorting(sa) | sa:SATdx } || || auction() || out-buffer() || ministry() auction_master() = auctioneer() || office() coop: Cldx -> Unit cooperative(i) = || {lift-truck(t) | t:TIdx } |] || { sort.group(s) | s:SIdx } || coop-mnanager() The above expresses an essence of a fisheries indusry. It is based on such essential mod- els that comprehensive theories of the man-made universe of fisheries (and agricultural) industries can be researched and applied. 80