Conceptual Modeling 33rd International Conference ER 2014 Atlanta GA USA October 27 29 2014 Proceedings 1st Edition Eric Yu

Conceptual Modeling 33rd International
Conference ER 2014 Atlanta GA USA

October 27 29 2014 Proceedings 1st
Edition Eric Yu
Visit to download the full and correct content document:
https://textbookfull.com/product/conceptual-modeling-33rd-international-conference-er
-2014-atlanta-ga-usa-october-27-29-2014-proceedings-1st-edition-eric-yu/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Intelligent Virtual Agents 14th International

Conference IVA 2014 Boston MA USA August 27 29 2014
Proceedings 1st Edition Timothy Bickmore
https://textbookfull.com/product/intelligent-virtual-agents-14th-
international-conference-iva-2014-boston-ma-usa-
august-27-29-2014-proceedings-1st-edition-timothy-bickmore/
Mobile Web Information Systems 11th International

Conference MobiWIS 2014 Barcelona Spain August 27 29
2014 Proceedings 1st Edition Irfan Awan
https://textbookfull.com/product/mobile-web-information-
systems-11th-international-conference-mobiwis-2014-barcelona-
spain-august-27-29-2014-proceedings-1st-edition-irfan-awan/
Case Based Reasoning Research and Development 22nd

International Conference ICCBR 2014 Cork Ireland
September 29 2014 October 1 2014 Proceedings 1st
Edition Luc Lamontagne
https://textbookfull.com/product/case-based-reasoning-research-
and-development-22nd-international-conference-iccbr-2014-cork-
ireland-september-29-2014-october-1-2014-proceedings-1st-edition-
luc-lamontagne/
Knowledge Engineering and the Semantic Web 5th

International Conference KESW 2014 Kazan Russia
September 29 October 1 2014 Proceedings 1st Edition
Pavel Klinov
https://textbookfull.com/product/knowledge-engineering-and-the-
semantic-web-5th-international-conference-kesw-2014-kazan-russia-
september-29-october-1-2014-proceedings-1st-edition-pavel-klinov/
Combinatorial Optimization and Applications 12th
International Conference COCOA 2018 Atlanta GA USA
December 15 17 2018 Proceedings Donghyun Kim
https://textbookfull.com/product/combinatorial-optimization-and-
applications-12th-international-conference-cocoa-2018-atlanta-ga-
usa-december-15-17-2018-proceedings-donghyun-kim/
System Analysis and Modeling Models and Reusability 8th

International Conference SAM 2014 Valencia Spain
September 29 30 2014 Proceedings 1st Edition Daniel
Amyot
https://textbookfull.com/product/system-analysis-and-modeling-
models-and-reusability-8th-international-conference-
sam-2014-valencia-spain-september-29-30-2014-proceedings-1st-
edition-daniel-amyot/
Computer Safety Reliability and Security 33rd

International Conference SAFECOMP 2014 Florence Italy
September 10 12 2014 Proceedings 1st Edition Andrea
Bondavalli
https://textbookfull.com/product/computer-safety-reliability-and-
security-33rd-international-conference-safecomp-2014-florence-
italy-september-10-12-2014-proceedings-1st-edition-andrea-
bondavalli/
Simulation Modeling and Programming for Autonomous

Robots 4th International Conference SIMPAR 2014 Bergamo
Italy October 20 23 2014 Proceedings 1st Edition Davide
Brugali
https://textbookfull.com/product/simulation-modeling-and-
programming-for-autonomous-robots-4th-international-conference-
simpar-2014-bergamo-italy-october-20-23-2014-proceedings-1st-
edition-davide-brugali/
HCI International 2014 Posters Extended Abstracts

International Conference HCI International 2014
Heraklion Crete Greece June 22 27 2014 Proceedings Part
II 1st Edition Constantine Stephanidis (Eds.)
https://textbookfull.com/product/hci-international-2014-posters-
extended-abstracts-international-conference-hci-
international-2014-heraklion-crete-greece-
Eric Yu
Gillian Dobbie
Matthias Jarke
Sandeep Purao (Eds.)
LNCS 8824
Conceptual Modeling
33rd International Conference, ER 2014
Atlanta, GA, USA, October 27–29, 2014
Proceedings
123
Lecture Notes in Computer Science 8824
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David Hutchison
Lancaster University, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Alfred Kobsa
University of California, Irvine, CA, USA
Friedemann Mattern
ETH Zurich, Switzerland
John C. Mitchell
Stanford University, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
Oscar Nierstrasz
University of Bern, Switzerland
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbruecken, Germany
Eric Yu Gillian Dobbie Matthias Jarke
Sandeep Purao (Eds.)
Conceptual Modeling
33rd International Conference, ER 2014
Atlanta, GA, USA, October 27-29, 2014
Proceedings
13
Volume Editors
Eric Yu
University of Toronto
Faculty of Information
Toronto, ON, M5S 3G6, Canada
E-mail: eric.yu@utoronto.ca
Gillian Dobbie
University of Auckland
Department of Computer Science
Auckland 1142, New Zealand
E-mail: gill@cs.auckland.ac.nz
Matthias Jarke
RWTH Aachen University
Lehrstuhl Informatik 5
52056 Aachen, Germany
E-mail: jarke@dbis.rwth-aachen.de
Sandeep Purao
Penn State University
University Park
PA 16802, USA
E-mail: spurao@ist.psu.edu
ISSN 0302-9743 e-ISSN 1611-3349

ISBN 978-3-319-12205-2 e-ISBN 978-3-319-12206-9
DOI 10.1007/978-3-319-12206-9
Springer Cham Heidelberg New York Dordrecht London
Library of Congress Control Number: 2014950401

LNCS Sublibrary: SL 3 – Information Systems and Application, incl. Internet/Web
and HCI
© Springer International Publishing Switzerland 2014
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and
executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication
or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location,
in ist current version, and permission for use must always be obtained from Springer. Permissions for use
may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution
under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or
omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.
Typesetting: Camera-ready by author, data conversion by Scientific Publishing Services, Chennai, India
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)
Preface
The 33rd edition of the International Conference on Conceptual Modeling (ER),

held this year in Atlanta, Georgia, USA, continued the long tradition of the
ER conference series as the leading international forum for presenting research
results and discussing emerging trends in conceptual modeling.
Since the seminal work of Peter Chen on the entity relationship (ER) model,
the field of conceptual modeling has had profound impact on many areas of
computing and information systems, including databases, software, business pro-
cesses, and Internet systems and services. Conceptual modeling consists of the
creation of foundational elements of intended information systems and applica-
tions that can serve as communication vehicles among stakeholders, blueprints
for developers, as well as long-term preservation of designer intent. Conceptual
models, therefore, serve a critical role in the design and deployment of new tech-
nologies and applications. As a set of languages, tools and techniques, conceptual
models and modeling provide a rich arena for researchers with immediate appli-
cations to practice.
This foundational role of conceptual models was evident in this year’s pro-
gram, which included research papers addressing many contemporary topics
such as big data, city informatics, policy compliance, enterprise architecture,
data-intensive ecosystems, and open source software strategies. In addition, the
research community continues to investigate fundamental concerns of conceptual
modeling such as pragmatics, deployment paths, and modeling primitives.
The call for papers this year included a focus on fundamental concerns as
well as contemporary application areas. The international research community
responded by submitting a total of 123 abstracts, which resulted in 80 full paper
submissions. With the use of an online review system, each paper was reviewed by
at least three reviewers. On the basis of these reviews, we selected 23 submissions
as full papers (an acceptance rate of 28.75%). The authors of a further 15 papers
were invited to present their papers based on the contemporary nature of these
research efforts and the promise for future impact. These papers are included as
short papers in the proceedings.
The technical program at the conference consisted of a number of sessions
covering the aforementioned papers. In addition, the program consisted of panels
related to modeling for extreme events, conceptual models related to security
concerns, and other fundamental concerns. There were two tutorials, one devoted
to meta-models and meta-execution models, and a second related to creating new
requirements modeling languages.
In addition to these, the conference included a number of workshops related
to contemporary application areas. Papers from these workshops and papers
describing research software demonstrations are included in a companion volume
to these proceedings.
VI Preface
Finally, three highly interesting keynote presentations addressed important

current topics in the theory and practice of conceptual modeling. Financial data
mining pioneer Vasant Dhar (New York University, USA) pointed out critical
issues in predictive modeling in the context of big data and data science, in-
cluding the quest for an additional set of skills that students and practitioners
in the field will need to acquire. Giancarlo Guizzardi (Federal University of Es-
pirito Santo, Brazil, and University of Trento, Italy) emphasized quality control
of conceptual modeling using real-world semantics. Antonio Furtado (Pontifi-
cal Catholic University of Rio de Janeiro, Brazil) emphasized the advantages of
a semiotic perspective as another important foundation of formally supported
database design and usage.
We would like to thank all those who helped put this program together. We
would especially like to thank the individuals who have spearheaded a number
of activities. This includes Workshop Chair Marta Indulska; Doctoral Consor-
tium Chair Carson Woo; Tutorial Chair Cecil Chua; Panel Chair Sham Navathe;
Demonstrations Chair Akhilesh Bajaj; and Educational Symposium Chair Lois
Delcambre. In addition, Matti Rossi helped us in publicizing the conference;
Jon Becker kept the website updated regularly. Antoni Olivé, ER Steering Com-
mittee Chair, and Veda Storey, ER Steering Committee Liaison, were always
generous with their time in answering questions and providing guidance. We
also thank Stuart Madnick and Colette Rolland, who provided valuable guid-
ance as honorary chairs. Finally, Richard Welke was instrumental in ensuring
that the logistics ran smoothly, as the local arrangements chair. Special thanks
to Greg O’Dillon, who made a major contribution to making everything happen
in and around the conference venue.
We are grateful for the support from all the sponsors of the conference. Major
support for the conference was provided by Penn State University and Singapore
Management University. We wish to express our appreciation for Georgia State
University as the local organizing partner.
We sincerely thank the members of the Program Committee and external
reviewers who provided thoughtful reviews on time. Their reviews provided
feedback as well as suggestions that have significantly improved the technical
program as well as each individual paper.
Most importantly, we thank the authors for submitting their work and for
participating in the conference.
July 2014 Eric Yu

Gill Dobbie
Matthias Jarke
Sandeep Purao
Conference Organization
Honorary Chairs
Stuart Madnick MIT, Boston, MA, USA
Colette Rolland Université Paris 1 Panthéon – Sorbonne, France
Conference Chair
Sandeep Purao Penn State University, State College, PA, USA
Technical Program Co-chairs

Eric Yu University of Toronto, Canada
Gill Dobbie University of Auckland, New Zealand
Matthias Jarke RWTH Aachen University and Fraunhofer FIT,
Germany
Workshop Chair
Marta Indulska University of Queensland, Australia
Publicity Chair
Matti Rossi Aalto University, Finland
Panel Chair
Sham Navathe Georgia Tech, USA
Tutorial Chair
Cecil Chua Auckland University, New Zealand
Poster / Demo Chair

Akhilesh Bajaj University of Tulsa, USA
VIII Conference Organization
PhD Symposium Chair

Carson Woo University of British Columbia, Canada
Local Organization Chair

Richard Welke Georgia State University, USA
Sponsorships Co-chairs
Sandeep Purao Penn State University, USA
Aditya Ghose University of Wollongong, Australia
Wolfgang Maass Universität des Saarlandes, Germany
Finance Chair
Anna Squicciarini Penn State University, State College, USA
Webmaster
Jon Becker Penn State University, State College, USA
Steering Committee Liaison

Veda Storey Georgia State University, USA
Program Committee
Jacky Akoka CNAM and TEM, France
Yuan An Drexel University, USA
Joao Araujo Universidade Nova de Lisboa, Portugal
Zhifeng Bao University of Singapore, Singapore
Sandro Bimonte IRSTEA, France
Shawn Bowers Gonzaga University, USA
Stephane Bressan National University of Singapore, Singapore
Stefano Ceri Politecnico di Milano, Italy
Roger Chiang University of Cincinnati, USA
Dickson Chiu University of Hong Kong, SAR China
Conference Organization IX
Isabel F. Cruz University of Illinois, USA

Alfredo Cuzzocrea ICAR-CNR and University of Calabria, Italy
Fabiano Dalpiaz Universiteit Utrecht, The Netherlands
Karen Davis University of Cincinnati, USA
Valeria De Antonellis University of Brescia, Italy
Lois Delcambre Portland State University, USA
Joerg Evermann Memorial University of Newfoundland, Canada
Xavier Franch UniversitatPolitècnica de Catalunya, Spain
Avigdor Gal Technion-Israel Institute of Technology, Israel
Aditya Ghose University of Wollongong, Australia
Paolo Giorgini University of Trento, Italy
Georg Grossmann University of Australia, Australia
Giancarlo Guizzardi Federal University of Espirito Santo, Brazil
Arantza Illarramendi Basque Country University, Spain
Hemant Jain University of Wisconsin-Milwaukee, USA
Manfred Jeusfeld Tilburg University, The Netherlands
Ivan Jureta University of Namur, Belgium
Dimitris Karagiannis University of Vienna, Austria
David Kensche Thinking Networks AG, Germany
Alberto Laender Federal University of Minas Gerais, Brazil
Sang-Goo Lee Seoul National University, Korea
Mong Li Lee National University of Singapore, Singapore
Dik Lun Lee Hong Kong University of Science and
Technology, SAR China
Wolfgang Lehner Technische Universität Dresden, Germany
Julio Cesar Leite PUC Rio de Janeiro, Brazil
Stephen Liddle Brigham Young University, USA
Tok Wang Ling National University of Singapore, Singapore
Sebastian Link The University of Auckland, New Zealand
Mengchi Liu Carleton University, Canada
Fred Lochovsky The Hong Kong University of Science and
Technology, SAR China
Pericles Loucopoulos Loughborough University, UK
Hui Ma Victoria University of Wellington, New Zealand
Heinrich C. Mayr Alpen-Adria-Universität Klagenfurt, Austria
Jan Mendling Wirtschafts Universität Wien, Austria
Haralambos Mouratidis University of Brighton, UK
John Mylopoulos University of Trento, Italy
Antoni Olivé Universitat Politècnica de Catalunya, Spain
Andreas L. Opdahl University of Bergen, Norway
José Palazzo M. De Oliveira Universidade Federal do Rio Grande do Sul,
Brazil
Jeffrey Parsons Memorial University of Newfoundland, Canada
X Conference Organization
Oscar Pastor Lopez University of Valencia, Spain

Zhiyong Peng Wuhan University, China
Barbara Pernici Politecnico di Milano, Italy
Geert Poels Ghent University, Belgium
Henderik Proper Public Research Centre Henri Tudor,
Luxembourg
Christoph Quix RWTH Aachen University, Germany
Jolita Ralyté University of Geneva, Switzerland
Sudha Ram University of Arizona, USA
Iris Reinhartz-Berger University of Haifa, Israel
Stefano Rizzi University of Bologna, Italy
Colette Rolland Université Paris 1 Panthéon – Sorbonne, France
Antonio Ruiz-Cortés University of Seville, Spain
Bernhard Rumpe RWTH Aachen University, Germany
Mehrdad Sabetzadeh University of Luxembourg, Luxembourg
Motoshi Saeki Tokyo Institute of Technology, Japan
Camille Salinesi Université Paris 1 Panthéon – Sorbonne, France
Peretz Shoval Ben-Gurion University, Israel
Il-Yeol Song Drexel University, USA
Steffen Staab University of Koblenz-Landau, Germany
Veda Storey Georgia State University, USA
Arnon Sturm Ben-Gurion University of the Negev,
Israel
David Taniar Monash University, Australia
Ernest Teniente Universitat Politècnica de Catalunya, Spain
James Terwilliger Microsoft Corporation, USA
Bernhard Thalheim University of Kiel, Germany
Juan Trujillo University of Alicante, Spain
Panos Vassiliadis University of Ioannina, Greece
Ramesh Venkataraman Indiana University, USA
Gerd Wagner Brandenburg University of Technology at
Cottbus, Germany
Barbara Weber University of Innsbruck, Austria
Roel Wieringa University of Twente, The Netherlands
Carson Woo University of British Columbia, Canada
Huayu Wu Agency for Science, Technology and Research,
Singapore
Esteban Zimányi Université Libre de Bruxelles, Belgium
Additional Reviewers
Bertossi, Leopoldo Cappiello, Cinzia

Bianchini, Devis De Kinderen, Sybren
Bjekovic, Marija Fernandez, Pablo
Bork, Domenik Guarino, Nicola
Conference Organization XI
Horst, Andreas Navarro Perez, Antonio

Kim, Hyun Joon Palmonari, Matteo
Kim, Meenv Chul Peña, Joaquı́n
Koehler, Henning Plataniotis, Georgios
Lee, Hanbit Resinas, Manuel
Ling, Yuan Sagi, Tomer
Liu, Mengwen Shin, Youhyun
Manousis, Petros Sugumaran, Vijayan
Marosin, Diana Visic, Niksa
Mascaro, Christopher Wu, Feng
Mate, Alejandro Yu, Ting
Melchiori, Michele Zang, Yizhou
Mir Seyed Nazari, Pedram Zhu, Yongjun
XII Conference Organization
Sponsors
Gold Sponsors
Silver Sponsors
Bronze Sponsors
Keynotes (Abstracts)
Data Science and Prediction
Vasant Dhar
Professor and Director

Center for Business Analytics
New York University
vdhar@stern.nyu.edu
Abstract. Data Science is the study of the generalizable extraction

of knowledge from data. A common epistemic requirement in assessing
whether new knowledge is actionable for decision making is its predic-
tive power, not just its ability to explain the past. The heterogeneity and
scale of data and diversity of analytical methods require data scientists
to have an integrated skill set, as well as a deep understanding of the
craft of problem formulation and the science required to engineer effec-
tive solutions. I shall talk about the key issues that arise in industrial
strength predictive modeling, including the implications for education in
this fast emerging field.
A Semiotic Approach to Conceptual Modelling
Antonio L. Furtado, Marco A. Casanova, and Simone D.J. Barbosa
Departmento de Informática
Pontifı́cia Universidade Católica do Rio de Janeiro (PUC-Rio)
Rio de Janeiro, Brazil
{furtado,casanova,simone}@inf.puc-rio.br
Abstract. The work on Conceptual Modelling performed by our group

at PUC-Rio is surveyed, covering four mutually dependent research top-
ics. Regarding databases as a component of information systems, we
extended the scope of the Entity-Relationship model, so as to encom-
pass facts, events and agents in a three-schemata specification method
employing a logic programming formalism. Next we proceeded to ren-
der the specifications executable, by utilizing backward-chaining plan-
ners to satisfy the agents’ goals through sequences of fact-modification
events. Thanks to the adoption of this plan-recognition / plan-generation
paradigm, it became possible to treat both business-oriented and fictional
narrative genres. To guide our conceptual modelling approach, we iden-
tified four semiotic relations, associated with the four master tropes that
have been claimed to provide a system to fully grasp the world concep-
tually.
Keywords: Entity-Relationship Model, Information Systems, Planning,

Logic Programming, Narrative Genres, Semiotics.
Ontological Patterns, Anti-Patterns and Pattern
Languages for Next-Generation Conceptual
Modeling
Giancarlo Guizzardi
Ontology and Conceptual Modeling Research Group (NEMO),

Computer Science Department,
Federal University of Espı́rito Santo (UFES), Vitória - ES, Brazil
gguizzardi@inf.ufes.br
Abstract. This paper addresses the complexity of conceptual modeling

in a scenario in which semantic interoperability requirements are increas-
ingly present. It elaborates on the need for developing sound ontological
foundations for conceptual modeling but also for developing complex-
ity management tools derived from these foundations. In particular, the
paper discusses three of these tools, namely, ontological patterns, onto-
logical anti-patterns and pattern languages.
Keywords: Conceptual Modeling, Formal Ontology, Patterns,

Anti-Patterns and Pattern Languages
Table of Contents
Keynotes
A Semiotic Approach to Conceptual Modelling . . . . . . . . . . . . . . . . . . . . . . 1
Antonio L. Furtado, Marco A. Casanova, and
Simone Diniz Junqueira Barbosa
Ontological Patterns, Anti-Patterns and Pattern Languages for

Next-Generation Conceptual Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Giancarlo Guizzardi
Data on the Web

A Computer-Guided Approach to Website Schema.org Design . . . . . . . . . 28
Albert Tort and Antoni Olivé
On Designing Archiving Policies for Evolving RDF Datasets

on the Web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Kostas Stefanidis, Ioannis Chrysakis, and Giorgos Flouris
Ontology-Based Spelling Suggestion for RDF Keyword Search . . . . . . . . . 57

Sheng Li, Junhu Wang, and Xin Wang
Unstructured Data
Schema-Independence in XML Keyword Search . . . . . . . . . . . . . . . . . . . . . . 71
Thuy Ngoc Le, Zhifeng Bao, and Tok Wang Ling
Mapping Heterogeneous XML Document Collections to Relational

Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Prudhvi Janga and Karen C. Davis
MKStream: An Efficient Algorithm for Processing Multiple Keyword

Queries over XML Streams (Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Evandrino G. Barros, Alberto H.F. Laender, Mirella M. Moro, and
Altigran S. da Silva
Uncertain and Incomplete Data

Cardinality Constraints for Uncertain Data . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Henning Koehler, Sebastian Link, Henri Prade, and Xiaofang Zhou
XX Table of Contents
TopCrowd – Efficient Crowd-enabled Top-k Retrieval on Incomplete

Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Christian Nieke, Ulrich Güntzer, and Wolf-Tilo Balke
Web Services Composition in the Presence of Uncertainty

(Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Soumaya Amdouni, Mahmoud Barhamgi, Djamal Benslimane,
Rim Faiz, and Kokou Yetongnon
Big Data, Graphs and Networks

Domain Ontology As Conceptual Model for Big Data Management:
Application in Biomedical Informatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Catherine Jayapandian, Chien-Hung Chen, Aman Dabir,
Samden Lhatoo, Guo-Qiang Zhang, and Satya S. Sahoo
Network Analytics ER Model – Towards a Conceptual View of Network

Analytics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Qing Wang
Model-Driven Design of Graph Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Roberto De Virgilio, Antonio Maccioni, and Riccardo Torlone
Privacy and Safety

Utility-Friendly Heterogenous Generalization in Privacy Preserving
Data Publishing (Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Xianmang He, Dong Li, Yanni Hao, and Huahui Chen
From Conceptual Models to Safety Assurance . . . . . . . . . . . . . . . . . . . . . . . 195

Yaping Luo, Mark van den Brand, Luc Engelen, and
Martijn Klabbers
Database Design
A New Approach for N-ary Relationships in Object Databases . . . . . . . . . 209
Jie Hu, Liu Chen, Shuang Qiu, and Mengchi Liu
Database Design for NoSQL Systems (Short paper) . . . . . . . . . . . . . . . . . . 223

Francesca Bugiotti, Luca Cabibbo, Paolo Atzeni, and
Riccardo Torlone
Fixing Up Non-executable Operations in UML/OCL Conceptual

Schemas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Xavier Oriol, Ernest Teniente, and Albert Tort
Table of Contents XXI
Generic Data Manipulation in a Mixed Global/Local Conceptual

Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Scott Britell, Lois M.L. Delcambre, and Paolo Atzeni
New Modeling Languages and Applications

Evaluating Modeling Languages: An Example from the Requirements
Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Jennifer Horkoff, Fatma Başak Aydemir, Feng-Lin Li, Tong Li, and
John Mylopoulos
Nòmos 3: Legal Compliance of Roles and Requirements . . . . . . . . . . . . . . . 275

Silvia Ingolfo, Ivan Jureta, Alberto Siena, Anna Perini, and
Angelo Susi
Towards an XBRL Ontology Extension for Management Accounting

(Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Barbara Livieri, Marco Zappatore, and Mario Bochicchio
Representing Hierarchical Relationships in INM (Short paper) . . . . . . . . . 297

Mengchi Liu, Jie Hu, Liu Chen, and Xuhui Li
Providing Foundation for User Feedback Concepts by Extending a

Communication Ontology (Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Itzel Morales-Ramirez, Anna Perini, and Renata Guizzardi
Towards a Conceptual Framework and Metamodel for Context-

Aware Personal Cross-Media Information Management Systems
(Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Sandra Trullemans and Beat Signer
Software Concepts and Strategies

Software as a Social Artifact: A Management and Evolution
Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Xiaowei Wang, Nicola Guarino, Giancarlo Guizzardi, and
John Mylopoulos
Modelling Risks in Open Source Software Component Selection . . . . . . . . 335

Alberto Siena, Mirko Morandini, and Angelo Susi
Modelling and Applying OSS Adoption Strategies . . . . . . . . . . . . . . . . . . . . 349

Lidia López, Dolors Costal, Claudia P. Ayala, Xavier Franch,
Ruediger Glott, and Kirsten Haaland
XXII Table of Contents
Patterns and Narratives

Detection, Simulation and Elimination of Semantic Anti-Patterns in
Ontology-Driven Conceptual Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
Giancarlo Guizzardi and Tiago Prince Sales
Recall of Concepts and Relationships Learned by Conceptual Models:

The Impact of Narratives, General-Purpose, and Pattern-Based
Conceptual Grammars (Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Wolfgang Maass and Veda C. Storey
Visual Maps for Data-Intensive Ecosystems (Short paper) . . . . . . . . . . . . . 385

Efthymia Kontogiannopoulou, Petros Manousis, and
Panos Vassiliadis
Data Management for Enterprise Architecture

A Framework for a Business Intelligence-Enabled Adaptive Enterprise
Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
Okhaide Akhigbe, Daniel Amyot, and Gregory Richards
Modeling Organizational Alignment (Short paper) . . . . . . . . . . . . . . . . . . . 407

Henrique Prado Sousa and Julio Cesar Sampaio do Prado Leite
Compliance with Multiple Regulations (Short paper) . . . . . . . . . . . . . . . . . 415

Sepideh Ghanavati, Llio Humphreys, Guido Boella, Luigi Di Caro,
Livio Robaldo, and Leendert van der Torre
CSRML4BI: A Goal-Oriented Requirements Approach for Collaborative

Business Intelligence (Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
Miguel A. Teruel, Roberto Tardı́o, Elena Navarro, Alejandro Maté,
Pascual González, Juan Trujillo, and Rafael Muñoz-Terol
Embracing Pragmatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

Marija Bjeković, Henderik A. Proper, and Jean-Sébastien Sottet
City and Urban Applications

From Needs to Services: Delivering Personalized Water Assurance
Services in Urban Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
Kala Fleming and Komminist Weldemariam
Modeling Claim-Making Process in Democratic Deliberation

(Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Ye Tian and Guoray Cai
Table of Contents XXIII
Creating Quantitative Goal Models: Governmental Experience

(Short paper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Okhaide Akhigbe, Mohammad Alhaj, Daniel Amyot,
Omar Badreddin, Edna Braun, Nick Cartwright,
Gregory Richards, and Gunter Mussbacher
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475

A Semiotic Approach to Conceptual Modelling
Antonio L. Furtado, Marco A. Casanova, and Simone Diniz Junqueira Barbosa
Departmento de Informática
Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)
Rio de Janeiro, Brazil
{furtado,casanova,simone}@inf.puc-rio.br
Abstract. The work on Conceptual Modelling performed by our group at PUC-

Rio is surveyed, covering four mutually dependent research topics. Regarding
databases as a component of information systems, we extended the scope of the
Entity-Relationship model, so as to encompass facts, events and agents in a
three-schemata specification method employing a logic programming formal-
ism. Next we proceeded to render the specifications executable, by utilizing
backward-chaining planners to satisfy the agents' goals through sequences of
fact-modification events. Thanks to the adoption of this plan-recognition /
plan-generation paradigm, it became possible to treat both business-oriented
and fictional narrative genres. To guide our conceptual modelling approach, we
identified four semiotic relations, associated with the four master tropes that
have been claimed to provide a system to fully grasp the world conceptually.
Keywords: Entity-Relationship Model, Information Systems, Planning, Logic

Programming, Narrative Genres, Semiotics.
1 Introduction
Our understanding of information systems comprises facts, events and agents. Every-
where the Entity-Relationship model is used. The existing entity instances and their
properties, i.e., their attributes and the relationships among them, are the facts that
characterize a state of the world. States are changed by the occurrence of events
caused by operations defined by pre-conditions and post-conditions that are in turn
expressed in terms of such facts. The event-producing operations are performed by
certain agents, in an attempt to satisfy their goals, once again expressed by facts. Ac-
cordingly, our specifications are divided into three schemas to introduce, respectively,
the classes of facts (static schema), events (dynamic schema) and agents (behavioural
schema).
It so happens that the pre-conditions to bring about an event may need to be first
fulfilled as part of the effect (post-conditions) of other events. This partial-order
dependence immediately suggests the recursive application of backward-chaining
plan-generators in order to find one or more sequences of operations (plans) able to
perform a transition from the current state to a state wherein the goals of an agent
hold. By using a logic programming notation to represent the three schemas and
E. Yu et al. (Eds.): ER 2014, LNCS 8824, pp. 1–12, 2014.

© Springer International Publishing Switzerland 2014
2 A.L. Furtado, M.A. Casanova, and S.D.J. Barbosa
having developed in Prolog a plan-generation algorithm, we gained the benefit of

executable specifications, enabling to simulate and helping to gradually produce a
running system. Moreover, plans previously generated by the algorithm, or originated
from regulations or from customary practice, can be converted into patterns and kept
in libraries from which they can be retrieved by a plan-recognition algorithm and
reused, after the necessary adaptations, to reach similar goals.
Besides business information systems domains, we soon realized that our specifi-
cation method was applicable to narrative genres in general, ranging from "serious"
applications, such as maintenance procedures for an oil company, to fictional sword-
and-dragon stories. Indeed plot composition can be conveniently achieved via interac-
tive plan-generation. Alternatively, narrative motifs can serve as patterns to be
retrieved from a library and combined to compose the plots, thus providing another
opportunity to utilize plan-recognition.
When specifying any system, and when using it as well, some guidelines should
be available. What properties are relevant to characterize an object? What events
should be observed? How do agents interact, either collaborating or competing? Is it
possible to attain modularity, by setting the focus to different degrees of detail?
Which integrity constraints should be enforced? We found that four semiotic relations
establish and delimit the information space, covering the need for helpful reasoning
principles to a comprehensive extent. The syntagmatic relations determine – employ-
ing semiotic terminology [53] – a horizontal axis expressing the notion of connec-
tivity between information components. Another notion is similarity expressed by
paradigmatic relations, along a vertical axis. A depth axis, expressing granularity,
results from meronymic relations. Finally, topological limits are imposed to this three-
dimensional space by antithetic relations, which express negation and opposition. The
four semiotic relations are associated with the so-called four master tropes (metonym,
metaphor, synecdoche, and irony) [9], thought to constitute "a system, indeed the
system, by which the mind comes to grasp the world conceptually in language" [23].
Our research work on conceptual modelling, motivated by the considerations
above, is briefly surveyed as follows. Section 2 refers to the separate but consistently
adjusted specification of facts, events and agents. Section 3 covers plan-generation
and plan-recognition. Section 4 addresses application domains and narrative genres.
Section 5, the longest one, discusses the semiotic relations. Section 6 presents the
concluding remarks.
2 Three-Schemata Specifications
We have been working with the conceptual modeling of information systems with a
database component, considering their static, dynamic and behavioral aspects.
The static aspect concerns what facts hold at some database state, conveniently
described in terms of the entity-relationship model.
The dynamic aspect corresponds to events that can produce state transitions.
Events result from the execution of operations, defined in a declarative style by
their pre-conditions and post-conditions, according to the STRIPS proposal [25].
A Semiotic Approach to Conceptual Modelling 3
Pre-conditions involve the presence or absence of facts, whereas post-conditions

comprise the sets of facts added or deleted as the effect of the operation. Adopting the
notion of abstract data types, implicit in object-oriented approaches, we require that
facts can only be modified through the execution of such operations, whose pre- and
post-conditions are adjusted so as to preserve all integrity constraints.
The behavioural aspect refers to the agents authorized to cause events by perform-
ing the operations. To model this aspect we mainly use goal-inference rules, which
indicate what facts should hold, or cease to hold, at a target state that an agent will be
motivated to bring about in view of a situation, again expressed in terms of facts hold-
ing or not holding, prevailing at the current state [22]. In order to reach the desired
target state, an agent would execute – or ask the authorized agents to execute – some
appropriate plan, composed of one or more pre-defined operations. As a further
development, we have started to look at agent profiles involving three kinds of per-
sonality factors, from which a decision-making process could operate: drives for the
emergence of goals from situations, attitudes for the choice of plans to achieve the
preferred goal, and emotions to decide whether or not to commit to the execution of
the chosen plan, depending on the expected emotional gain when passing from the
current to the target state [4,40]. As an inducement to revise individual decisions, we
included competition and collaboration interferences, as prescribed for multi-agent
contexts [54].
3 The Plan-Recognition / Plan-Generation Paradigm
The three aspects treated in the preceding section were integrated through the applica-
tion of a plan-recognition / plan-generation paradigm [35,38].
In order to make our conceptual specifications executable [32], we created an en-
vironment where entity and relationship classes, operations, and goal-inference rules
and agent profiles are all represented as Prolog clauses. Also written in Prolog, algo-
rithms were provided for planning and for the simulated execution of the generated
plans [17,18,19,29]. Moreover it was noted that simulation can become a useful re-
source to support learning or training [20].
The plan-recognition side of the paradigm is relevant, after the system has been
made operational, as a means to extend conventional query facilities towards truly
cooperative responses. Cooperation, as discussed in section 5, is most effective when
one can detect what the user is trying to accomplish. The plan-recognition algorithm,
which we adapted from [46], matches a few observed actions of the user against a
library of previously recorded typical plans. The observed actions can be taken from
the execution log, which is updated whenever each operation of a transaction of the
user's initiative is executed. As we explained in [39], the library of typical plans, in
turn, can be constructed by inspecting the log and extracting and filtering sequences
of executed operations, whereby the transition indicated in some goal-inference rule
has been achieved.
4 Application Domains and Narrative Genres
Treating databases as a component of information systems encompassing facts, events

and agents permits a shift from a purely descriptive to a narrative context [27]. Indeed
in a workshop devoted to the application of natural languages to information systems,
we showed how to generate template-based natural language text, by inspecting the
plot-structured execution log and analyzing it against our three-level conceptual
schemas [37]. It is therefore not surprising that all the discussion in this section ap-
plies in essentially the same way to literary genres [15,28,36], whenever the fictional
events can be equally attributed to a pre-defined repertoire of operations performed by
agents (cf. the functions and dramatis personae in [50]). Recognizing that literary
genres ruled by identifiable conventions can thus be treated as one more kind of ap-
plication domain, we have adopted plan-based plot composition, coupled with several
dramatization techniques and visual media, within an ongoing digital storytelling
project [16,21].
The application of the plan-recognition / plan-generation paradigm to the narrative
domain [28] was presented at the XIX Brazilian Symposium on Data Bases as an
invited talk, on which occasion the author received a prize from the Brazilian Com-
puter Society, acknowledging his contributions to database research.
Having started in the fictional genres with folktale sword-and-dragon stories, we
have recently moved to genres where, besides action events, communicative events
must be specified and play a decisive role, such as detective stories [5].
5 Semiotic Completeness
Based on studies [9,12,51,55] asserting the completeness as reasoning processes of

the so-called four master tropes – metonymy, metaphor, irony and synecdoche, we
identified four types of semiotic relations that can exist not only between facts, but
also between events and between agents, which we denominated, respectively, syn-
tagmatic, paradigmatic, antithetic and meronymic relations. Informally speaking,
syntagmatic relations refer to connectivity, paradigmatic relations to similarity and
analogy, antithetic relations to negation, and meronymic relations to hierarchy.
Meronymy was, curiously, treated in our very first participation in Entity-
Relationship events [52], when we proposed to include semantic is-a and part-of hier-
archies into the ER model. Not much later we learned about the seminal contribution
of [56], where six types of part-of were distinguished.
The paradigmatic relations were the next to attract our attention. In a SIGPLAN
Notices paper [26], belonging to logic programming rather than to the database area,
we argued that a powerful kind of reasoning by analogy is provided by combining
unification with most specific generalization. We presented a revised version of an
existing algorithm to compute the most specific generalization of terms, which cor-
rectly decides whether or not new variables should be introduced in each case. We
also provided programs to perform unification and most specific generalization over
frames, a data structure which would be of major importance for the practical applica-
tion of our ideas, as will be repeatedly stressed in the sequel.
Our paper presented at SBBD 2007 [2] can be regarded as a first attempt to deal
with paradigmatic relations in the context of databases. The motivating problem was
that databases, particularly when storing heterogeneous, sparse semi-structured data,
tend to provide incomplete information and information which is difficult to catego-
rize. The paper first considers how to classify entity instances as members of entity
classes organized in a lattice-like generalization/specialization hierarchy. Then, it
describes how the frame representation employed for instances and classes, as well as
the closeness criterion involved in the classification method, favors the practical use
of similarity and analogy, where similarity refers to instances within the same class,
and analogy involves different classes. Finally, the paper argues that similarity and
analogy facilitate querying semi-structured data.
A more in-depth investigation of classification methods based on frames was the
object of a more recent work [48]. In fact, the problem of data classification goes back
to the definition of taxonomies covering knowledge areas. With the advent of the
Web, the amount of data available increased several orders of magnitude, making
manual data classification impossible. The paper presents a tool to automatically clas-
sify semi-structured data, represented by frames, without any previous knowledge
about structured classes. The tool uses a variation of the K-Medoid algorithm and
organizes a set of frames into classes, structured as a strict hierarchy.
The next step, still focusing on paradigmatic relations and the corresponding
trope, metaphor, was to promote a reuse strategy, whereby new conceptual specifica-
tions might be partly derived from previous ones. A paper along this line was pre-
sented at CIKM [8]. Metaphor is not merely a rhetorical device, characteristic of
language alone, but rather a fundamental feature of the human conceptual system.
A metaphor is understood by finding an analogy mapping between two domains. The
paper argued that analogy mappings facilitate conceptual modeling by allowing the
designer to reinterpret fragments of familiar conceptual models in other contexts.
The contributions of the paper were expressed within the tradition of the ER model,
the Description Logic framework and as extensions of the OWL.
This reuse strategy was further examined in [7,30]. These papers argued in favor
of a database conceptual schema and Semantic Web ontology design discipline that
explores analogy mappings to reuse the structure and integrity constraints of concep-
tual models, stored in a repository. We presupposed that a team of expert conceptual
designers would build a standard repository of source conceptual models, which less
experienced designers would use to create new target conceptual models in other
domains. The target models will then borrow the structure and the integrity con-
straints from the source models by analogy. The concepts were once again expressed
in the contexts of Description Logic, the RDF model and OWL to reinforce the basic
principles and explore additional questions, such as the consistency of the target
model.
Reusing a conceptual schema is of course a multi-phase process. After finding a
suitable source schema, adaptations will often be needed in view of conflicts with the
target schema being designed. The notion of blending [24] was exploited for this
objective in [10]. To support the generation of database schemas of information

systems, the paper proposed a five-step design process that explores the notions of
generic and blended spaces and favors the reuse of predefined schemas. The use of
generic and blended spaces is essential to achieve the passage from the source space
into the target space in such a way that differences and conflicts can be detected and,
whenever possible, conciliated. The convenience of working with multiple source
schemas to cover distinct aspects of a target schema, as well as the possibility of cre-
ating schemas at the generic and blended spaces, was also considered. Notice that, as
we would indicate more explicitly in later articles, the presence of conflicts already
suggests the need to deal with antithetic relations.
As mentioned before, since our already referred SBBD paper [2], we have been
using frames and frame-sets as a more flexible data structure than relational tuples
and tables. At the 27th ER Conference [33], as we proceeded to show how to extend
the reuse strategy to the design of dynamic schemas, we employed plots, also defined
as a frame-like data structure. A plot is a partially ordered set of events. Plot analysis
is a relevant source of knowledge about the agents’ behavior when accessing data
stored in the database. It relies on logical logs, which register the actions of individual
agents. The paper proposed techniques to analyze and reuse plots based on the con-
cepts of similarity and analogy. The concept of similarity was applied to organize
plots as a library and to explore the reuse of plots in the same domain. By contrast, the
concept of analogy helps reuse plots across different domains. The techniques pro-
posed in the paper find applications in areas such as digital storytelling and emer-
gency response information system, as well as some traditional business applications.
Our first study wherein all four semiotic relations were explicitly discussed was
indeed presented at a digital storytelling conference, namely SBGames [14]. In that
paper, the process of plot composition in the context of interactive storytelling was
considered under a fourfold perspective, in view of syntagmatic, paradigmatic, anti-
thetic and meronymic relations between the constituent events. These relations were
then shown to be associated with the four major tropes of semiotic research. A con-
ceptual model and set of facilities for interactive plot composition and adaptation
dealing with the four relations was described. To accommodate antithetic relations,
corresponding to the irony trope, our plan-based approach leaves room for the un-
planned. A simple storyboarding prototype tool has been implemented to conduct
experiments. In another paper [4], already mentioned in Section 2, we utilized the
semiotic relations to characterize classes of characters (agents, in the context of busi-
ness information systems) according to their mutually interfering behavior in deci-
sion-making processes.
As remarked earlier, frames and plots became increasingly important to our re-
search projects. The ER model is arguably today's most widely accepted basis for the
conceptual specification of information systems. A further common practice is to use
the relational model at an intermediate logical stage, in order to adequately prepare
for physical implementation. Although the relational model still works well in con-
texts relying on standard databases, it imposes certain restrictions, not inherent in ER
specifications, which make it less suitable in Web environments. Our 28th ER Con-
ference invited paper [34], mentioned at the end of Section 3, recommends frames as
an alternative to move from ER specifications to logical stage modeling, and treats

frames as an abstract data type equipped with a Frame Manipulation Algebra. It is
argued that frames, with a long tradition in AI applications, are able to accommodate
the irregularities of semi-structured data, and that frame-sets generalize relational
tables, allowing to drop the strict homogeneity requirement. The paper includes ex-
amples to help describe the use of the operators.
Likewise, a Plot Manipulation Algebra was proposed to handle plots in an ICEC
conference [45]. The seven basic operators, equally named in both the Frame Manipu-
lation Algebra and in the Plot Manipulation Algebra, and working respectively on
frames and plots, were introduced in view of the four fundamental semiotic relations,
as indicated below:
• syntagmatic relations - product, projection
• paradigmatic relations - union, selection
• antithetic relations - difference
• meronymic relations - combination, factoring
The operators in the first three lines above encompass the equivalent to the five
basic operators of Codd's relational algebra (product, projection, union, selection,
difference). The additional two operators (combination, factoring) handle the hierar-
chical structures induced by the meronymic relations, a notion that would correspond
to non-first-normal form (NF2) relations in the relational model (cf. our algebra of
quotient relations [41]). Thus, it seems fair to claim that our algebras are semiotically
complete, a notion that covers an ampler scope than that of Codd's relational algebra.
Prototype logic-programming tools have been developed to experiment with the
Frame Manipulation Algebra and the Plot Manipulation Algebra.
The pragmatic aspects of information systems constitute the main thrust of our
present work, strongly influenced by the fundamental semiotic concepts exposed in
this section. At the Second Workshop of the Brazilian Institute for Web Science Re-
search [3], we argued for this orientation, which becomes increasingly relevant with
the transition from the closed world of the old proprietary databases to the open world
of the Web. Our view of information systems recognizes that, in order to serve as a
basis for an effective communication process, their conceptual specification is compa-
rable to the definition of a specialized language. Accordingly, it must pass through
four levels: lexical, syntactic, semantic, and pragmatic [47]. At the semantic level, the
correspondence between the stored data and real world facts is considered, but to
design systems of practical usefulness, one still needs to investigate what purposes
they will serve, which falls in the scope of the pragmatic level.
This pragmatic orientation, as we soon realized, is fully consistent with our con-
ceptual design method that, as gradually exposed in the preceding sections, encom-
passes not only facts, but also events and agents. Motivated by their goals, defined
in terms of database facts, agents try to cause the occurrence of events whereby a
database state is reached where the goals are satisfied. And our plan-recognition /
plan-generation paradigm puts together all these aspects and leads to executable
specifications, which allow simulation experiments to effectively test the usability of
the proposed conceptual design.
6 Concluding Remarks
An early proposal on the subject of database modelling [1] introduced an architecture

that puzzled both theoreticians and practitioners. What could be a "conceptual
schema"? None of the existing models seemed to offer an adequate basis to formulate
what was intended, namely the semantic contents of the stored data. As we all know,
this gap was appropriately filled by the Entity-Relationship model. Recognizing the
wise orientation taken by the model of describing the application domains in their
own language, our group proceeded to extend this direct and highly intuitive way to
characterize factual information to events and agents.
In the present time, the Web gives access to a continuously growing number of in-
formation sources and once again the word "navigation" is employed, no longer in the
sense of traversing an intricate network of physical pointers, but to designate the
novel opportunities opened by connectivity across linked Web pages. We are con-
vinced that detecting semiotic relations helps to guide navigation, not only exploring
connectivity but enriching the quest for information with similarity and inter-domain
analogy, allowing to zoom in and out to alternate between summaries and details, and
limiting excessive recall, in favour of precision, by negative directives purporting to
exclude irrelevant responses. With the objective of meeting the Semantic Web stan-
dards, some new formalisms have been proposed, but they clearly keep supporting the
conceptual modelling principles. Peter Chen himself has endorsed the statement that
"…RDF can be viewed as a member of the Entity-Relationship model family" [13].
Another contribution of the ANSI/X3/SPARC report that will continue to receive
close attention in our research project is the identification of external schemas,
whereby the participation of the different users should be duly taken into considera-
tion. By making the external schemas branch from the conceptual schema, the report
implicitly imposes, as a consequence, that they cannot be simply confused with views
extracted from relational tables. Conceptual modelling should first be applied to their
specification, which extends in one more direction the scope of the Entity-
Relationship model family.
A crucial semiotic notion applies whenever users, as human – rather than suppos-
edly tightly-controlled software agents – are concerned: to the signifier-signified cor-
respondence between an object and its representation [53], an interpretant [49] must
be interposed as a third component to stress that the correspondence is subject to each
person's understanding, which may be faulty or incomplete. Conceptual modelling at
the individual users' level must then deal with beliefs, as a correct or incorrect render-
ing of facts. When specifying an information system meeting, not only semantic, but
also pragmatic requirements, serious efforts should be invested in the design of ade-
quate interfaces that, as much as possible, avoid misconceptions and misconstruals
[42,43], and seek to identify the users' goals and plans [11], to maximize their satis-
faction while pursuing activities in consonance with ethical conduct and the adopted
procedural norms.
In Web environments, the possibility to tackle multiple sources raises to an espe-
cially critical level the problem of dealing with conflicting information. Examining
the process of communication as described in [44], showing a sender in the act of
Another random document with
no related content on Scribd:
Bureau of Mines and the War Department, put the
last ounce of ‘pep’ into the organization by asking us
to build 20,000 gas masks for shipment overseas.
20,000 masks did not seem like a very large order.
We did not fully appreciate all the conditions which a
war gas mask had to encounter, so we readily and
willingly accepted the order. Then began a struggle
with can manufacturers, buckle makers,
manufacturers of straps, rubber facepieces,
eyepieces, knapsacks, etc. The country was
canvassed from the Atlantic Coast to the Mississippi
River for manufacturers who could turn out the
different parts acceptably and in a hurry.
“Charcoal was made from red cedar by the Day
Chemical Co. of Westline, Pennsylvania; soda-lime
permanganate was manufactured by the General
Chemical Company; knapsacks by the Simmons
Hardware Company in St. Louis; facepieces by the
Goodrich and Goodyear Rubber Companies at
Akron; canisters by the American Can Company;
and the assembly made at one of the plants of the
American Can Company in Long Island City.
“The writer cannot recall all the doubts, fears,
optimism, and enthusiasm felt in turn by different
members of the organization during the fabrication
of those first 20,000 masks. We were performing an
important task for the War Department. Night
became day. Dewey, Lewis, Henderson, Gibbs, and
others stepped from one train to another, and we
used the telephone between Washington and St.
Louis or Boston as freely as we used the local
Washington telephone.
“We thought we could improve on the English
box respirator on various points. We made the
canister larger, and have been glad ever since that
we did. We thought the English mouthpiece was too
flexible and too small, and made ours stiff and
larger, and were sorry we made the change. We
tested the fillings against chlorine, phosgene,
prussic acid, etc., and had a canister that was all
that was desired for absorbing these gases. But,
alas, we did not know that chloropicrin was destined
to be one of the most important war gases used by
the various belligerents. Further, it was not fully
appreciated that the rubberized cloth used in
making the facepiece had to be highly impermeable
against gases, that hardness as much as anything
else was desired in the make-up of the soda-lime
granules in order to withstand rough jolting so that
the fines would not clog the canister, and raise the
resistance to breathing to a prohibitive figure.
Neither was it appreciated at that time by any of the
allies, that the gas mask really should be a be a
fighting instrument, one that men could work hard
in, run in, and wear for hours, without too serious
discomfort.
“The first 20,000 masks sent over to England
were completed by the Research Division in record
time. As compared with the French masks, they
were far superior, giving greater protection against
chlorine, phosgene, superpalite, prussic acid, xylyl
bromide, etc. The French mask was of the cloth
type, conforming to the face, and consisting of
twenty layers of cheesecloth impregnated with
sodium phenate and hexamethylenetetramine.
Chloropicrin went through this like a shot. Just
before the masks were sent abroad, we received
disturbing rumors of the contemplated use of large
quantities of chloropicrin. The French, apparently,
had no intention of changing the design of their
mask, and did not do so for months to come. We
therefore released the masks, they were sent
abroad, and an anxious research group on this side
of the water waited expectantly for the verdict. It
came. A brief cablegram told us what our English
cousins thought of us. It was a subject they had
been wrestling with for two years and a half. They
had had battlefield experience; they had gone
through the grief of developing poor masks into
better ones, knew the story better than we did, and
after a thorough test ‘hammered’ the American
design unmercifully.
“This experience put the Research Division on
its mettle. Our first attempt had given us the
necessary preliminary experience; cablegrams and
reports traveled back and forth; an expert or two
eventually came to this country from England in
response to previous appeals for assistance, and
we turned with adequate information to the
development of a real mask.”
The story of mustard gas is given later. It probably occupied more

time and thought on the part of the Research Division, as well as that
of Edgewood Arsenal and the Development Division, than any other
gas.
Diphenylchloroarsine led to the preparation of a series of arsenic
compounds, some more easily prepared and more or less effective.
Cyanogen chloride and cyanogen bromide, reported by the
Italians as having been used by the Germans, were extensively
studied.
The Inorganic Section was early interested in special incendiary
materials which were developed for bombs, shells, darts and
grenades, and which were later taken over by the Pyrotechnic
Section, and finally adopted by the Ordnance Department.
In discussing the work one can very well start with the Offense
Section. This Section had two aims in view always, to develop
methods of making the gases used by the Germans more
economically than they were making them, and to develop better
gases if possible. When we entered the war, chlorine, phosgene and
chloropicrin were the lethal gases used, while bromoacetone and
xylyl bromide were the lachrymators. It was not a difficult matter to
prepare these. But the introduction of mustard gas in the summer of
1917 and of diphenylchloroarsine in the autumn of the same year,
not only made our chemists ponder over a manufacturing method,
but also so revised our notions of warfare that the possibility of using
other substances created the need for extensive research. The
development of bromobenzylcyanide by the French likewise opened
a new field among lachrymatory substances.
Colored rockets and smokes were developed for the Navy and
Army. The smoke box was also studied but the work was taken over
by the Pyrotechnic Section.
A large amount of pure inorganic research on arsine and
arsenides, fluorine, hydrofluoric acid and fluorides, cyanides,
cyanogen sulfide and nitrogen tetroxide was carried out, sometimes
successfully and at other times with little or no success.
The Analytical Section not only carried out all routine analyses
but developed methods for many new gases.
The Offense Section worked in very close contact with the Small
Scale Manufacturing Section (Chemical Production Section). Often it
happened that a method, apparently successful in the laboratory,
was of no value in the plant. Small scale plants were developed for
mustard gas, hydrocyanic acid, cyanogen chloride, arsenic
trichloride, arsenic trifluoride, magnesium arsenide, superpalite and
bromobenzylcyanide.
The Chemical Defense Section, organized January, 1918, was
occupied with problems relating to protection, such as charcoal,
soda-lime, and special absorbents, eyepieces, smoke filters,
efficiency of absorbents, and special work with mustard gas.
Charcoal demanded extensive research. Raw materials required
a world-wide search, carbonizing methods had to be developed, and
impregnating agents were thoroughly studied. This story is told in
Chapter XIII.
Soda-lime was likewise a difficult problem. Starting with the
British formula, the influence of the various factors was studied and a
balance between a number of desirable qualities, absorptive activity,
capacity, hardness, resistance to abrasion, chemical stability, etc.,
obtained. The final product consisted of a mixture of lime, cement,
kieselguhr, sodium permanganate and sodium hydroxide.
Equally valuable work was performed in the perfection of two
carbon monoxide absorbents for the Navy. The better of these
consisted of a mixture of suitably prepared oxides which acts
catalytically under certain conditions, and causes the carbon
monoxide to react with the oxygen of the air. Since there are color
changes connected with the iodine pentoxide reaction (the first
absorbent) it has been possible to develop this so as to serve as a
very sensitive detector for the presence of carbon monoxide in air.
While the question of smoke filters was so important that it
occupied the attention of several Sections, the Defense Section
developed, as a part of its work, a standard method of testing and
comparing filters, and did a great deal of work on the preparation of
paper for this purpose.
Various problems related to mustard gas were also studied. The
question of a protective ointment was solved as successfully as
possible under the circumstances, but was dropped when it
appeared doubtful if under battlefield conditions of concentration and
length of exposure, any ointment offered sufficient protection to pay
for the trouble of applying it. The removal of mustard gas from
clothing was investigated, especially by the accelerating effect of
turkey red oil. Another phase of the work concerned the destruction
of mustard gas on the ground, while a fourth phase related to the
persistency of mustard (and other gases) on the field of battle.
The Gas Mask Research Section concerned itself largely with
developing methods of testing canisters and with routine tests. When
one considers the number of gases studied experimentally, the large
number of experimental canisters developed, all of which were
tested against two or more gases, and further that the Section
assisted in the control of the production at Long Island City, it is seen
that this was no small job. In addition, the effect of various
conditions, such as temperature, humidity, ageing, size of particles,
were studied in their relation to the life of absorbents and canisters.
Man tests and mechanical tests will be discussed in a later chapter.
Other studies were concerned with weathering tests of gas mask
fabrics, mustard gas detector, and covering for dugout entrances
(dugout blankets), which were impregnated with a mixture of mineral
and vegetable oils. In studying the course of gases through a
canister the “wave front” method was of great value in detecting
defects in canister design and filling.
The Pyrotechnic Section was composed of a number of units,
each with its own problem. The gas shell was studied, with special
reference to the stability of gases and toxic solids, both on storage
and on detonation. Extensive work was carried out on smoke
screens—a Navy funnel, an Army portable smoke apparatus, using
silicon tetrachloride, a grenade, a Livens, and various shell being
developed for that purpose. The smoke screen was adapted to the
tank and the airplane as well as to the funnel of a ship. Several types
of incendiary bombs and darts were perfected. The liquid fire gun
was studied but the results were never utilized because of the
abandonment as useless of that form of warfare. Various forms of
signal lights, flares, rockets and colored smokes were studied and in
most cases specifications were written. Extensive studies were also
carried out on gas shell linings, from which a lead and an enamel
lining were evolved. Many physical properties of war gases and their
mixtures were determined.
The Dispersoid Section studied the production of smokes or
mists from various solid and liquid substances. Apparatus were
developed to study the concentration of smoke clouds and their rate
of settling. The efficiency of various filters and canisters was
determined, and among other things, a new smoke candle was
perfected.
Mechanical research at first was related to design and
construction of a canister and mask, based on the English type.
During the latter part of 1917 the Tissot type of mask was studied
and then turned over to the Gas Defense Division. A Navy Head
Mask and canister was perfected. The horse mask was developed
along the lines of the British type, and also a dog mask of the same
general nature. Horse boots were also constructed, though they
never were used at the front. Many Ordnance and Pyrotechnic
problems were also successfully completed, not the least of which
was a noiseless gas cylinder. This section developed the first special
poison gas suit, composed of an oilcloth suit, a mask and helmet and
a special canister.
The Manufacturing Development Section had general charge of
the defense problems, and really acted as an emergency section,
filling in as occasion demanded. They developed mustard gas
clothing and a horse mask. They constructed a hydrogen plant at
Langley Field, assisted in solving the difficulties relating to Batchite
charcoal at Springfield, Mass., and co-operated in the study of paper
and felt as filtering materials for smokes. Towards the close of the
war the Section was interested in the application of the gas mask to
the industries.
The Physiological work is discussed under the Medical Division.
The Editorial Section received reports from all the other Sections,
from which a semi-monthly report was written, and distributed to
authorized representatives of the Army and Navy and to our Allies.
Reports were also received from abroad and the information thus
received was made available to the Research Division. As the
number of reports increased the work was collected together into
monographs on the various war gases, absorbents, smokes, etc.
After the signing of the armistice these were revised and increased
in number, so that about fifty were finally turned over to the Director
of the Chemical Warfare Service.
Gas Defense Division

The story of the Gas Defense Division is largely the story of the
gas mask. Colonel (then Mr.) Bradley Dewey was in charge of the
“first twenty thousand.” Soon after that work was undertaken, he was
commissioned Major in the Gas Defense Division of the Sanitary
Corps and was placed in charge of the entire manufacturing
program. The work of the Division included the development and
manufacture as well as the testing and inspection of gas masks, and
other defense equipment. The magnitude of the work is seen from
the following record of production: 5,692,000 completed gas masks,
3,614,925 of which were produced at the Long Island City Plant,
while the remainder were assembled at the Hero Manufacturing
Company’s Plant at Philadelphia, 377,881 horse masks, 191,388
dugout blankets, 2,450 protective suits and 1,773 pairs of gloves,
1,246 tons of protective ointment, 45,906 gas warning signals
(largely hand horns), 50,549 trench fans and many oxygen
inhalators.
Fig. 5.—The Defective Gas Mask.

Successfully used by the Gas Defense Division to stimulate care
in every part of the operation of the manufacture of Gas Masks.]
The story of the “first twenty thousand” has already been told on
page 43. That these masks were far from satisfactory is no reflection
upon the men who made them. Even with the standard design of the
British as a pattern, it was impossible to attain all the knowledge
concerning gas masks in two months. The experience gained in this
struggle enabled the Army to take up the manufacture of gas masks,
in July, 1917, with a more complete realization of the seriousness of
the task. The masks were not lost, either, for they were sent to the
various camps as training masks and served a very useful purpose.
The first order after this was for 1,100,000 masks, to be
completed within a year from date. For this production there was
authorized one major, two captains, and ten lieutenants. How little
the problem was understood is evident when we realize that in the
end there were 12,000 employees in the Gas Defense Plant at Long
Island City, N. Y. The first attempts were to secure these through
existing concerns. The Hero Manufacturing Company of Philadelphia
undertook the work and carried on certain portions of it all through
the War. Experience soon showed, however, that because of the
necessity for extreme care in the manufacture and inspection of the
mask, the ordinary commercial organization was not adapted to
carry on their manufacture on the scale necessitated by the Army
program. Consequently, on Nov. 21, 1917, the Secretary of War
authorized the establishment of a government operated plant, and
experienced officials were drawn from New York, Chicago, Boston
and other manufacturing centers to carry on the work. Buildings in
Long Island City, not far from the chemical plant (charcoal and soda
lime) at Astoria, were taken over by the officers of the Gas Defense
Service, until in July, 1918, five large buildings were occupied,
having a total floor space of 1,000,000 square feet (23 acres). The
organization grew from the original thirteen officers until it included
some 12,000 employees of whom about 8,500 were women.
Because of the care required in all the work, attempt was made to
secure, as far as possible, those who had relatives with the A. E. F.
The thought was that their personal interest in the work would result
in greater care in manufacture and inspection. The personnel was
unique in that the authority was apparently divided between civilian
and military, but there was no friction because of this. The efficiency
of the entire organization is shown by the fact that the masks
manufactured at Long Island City cost fifty cents less per mask than
those manufactured under contract.
The first actual shipment (overseas) of box respirators was made
from the Gas Defense Plant on March 4, 1918. From this date the
production increased by leaps and bounds. As mentioned above,
between this date and November 26, when the last mask was
manufactured, 3,146,413 masks of the box respirator type were
passed through final inspection in the plant. The greatest daily
production, 43,926 masks, was reached on October 26, 1918. The
process of manufacture will be discussed under the chapter on the
Gas Mask.
During the last half of 1918 the Kops Tissot mask was
manufactured. This mask had been perfected during the months
preceding August, 1918, when its manufacture was started.
Considerable difficulty was encountered in its production, but the first
mask was completed on September 14, and between that time and
the Armistice, 189,603 masks of this type had been manufactured.
Along with this manufacturing development went the building up
of an elaborate procurement force charged with the responsibility of
providing parts to be assembled at the Gas Defense Plant and at the
Hero Manufacturing Company. This Section faced a hard and
intricate task, but, though there were instances where the shortage
of parts temporarily caused a slowing down of production, these
were remarkably rare. Not only had the parts to be standardized, and
specifications written, but a field inspection force had to be trained in
order that the finished parts might be suitable for the final assembly
plant. The problem was further complicated by the fact that the
design was constantly changing, as improvement followed
improvement. Officers, trained in inspection in a day, were sent out
to train inspectors in the industrial centers.
In February, 1918, shortly before the German drive commenced,
requisitions were received for sample lots of oiled mittens and oiled
union suits as protection against mustard gas. These were prepared
in quantity and sent to the front, as was also a considerable amount
of chloride of lime for neutralizing the mustard gas in the field.
Another phase of the work consisted of the Field Testing Section,
which was organized to provide field testing conditions for the
regular product and for the development organization. Later there
were added a preliminary course of training for officers for overseas
duty in chemical warfare, the military training of the Gas Defense
officers located in and near New York and the training of boat crews
engaged in carrying offensive gas supplies. The Field Testing
Section rendered valuable service in pointing out weaknesses of
designs as developments took place and especially those
uncomfortable features of the masks which were apparent only
through long wear. During the course of this work the section built a
complete trench system in the Pennsylvania Railroad yards with an
elaborate dugout, the equal of any of the famous German quarters
on the Western front.
The chapters on Charcoal, Soda-Lime and the Gas Mask must
be read in this connection to gain an idea of the work carried out by
this Division. It is summed up in the statement that American soldiers
were provided with equipment which neutralized the best effects of
German chemical knowledge as evidenced by the offensive methods
and materials employed.
The organization of the Gas Defense Division, as of Nov. 11,
1918, was as follows:
Colonel Bradley Dewey Officer in Charge
Lieut. Col. A. L. Besse Asst. Officer in Charge
Major M. L. Emerson Administration Section
Major H. P. Schuit Comptrolling Section
Mr. R. Skemp Procurement Section
Major C. R. Johnson Technical Director
Capt. K. Atterbury Field Testing Section
Major J. C. Woodruff Chemical Manufacturing and
Development
Mr. R. R. Richardson Manager, Gas Defense Plant
Capt. H. P. Scott Officer in Charge,
Hero Manufacturing Co.
Major L. W. Cottman Engineering Branch
Major T. L. Wheeler Chemical Development
Major I. W. Wilson Astoria Branch
Capt. W. E. Brophy San Francisco Branch
Lt. E. J. Noble Cleveland Branch
Lt. L. Merrill Springfield Branch
Edgewood Arsenal
The Ordnance Department, in making plans for a shell filling
plant, thought to interest existing chemical firms in the manufacture
of the required toxic materials. As plans developed, however,
difficulties arose in carrying out this program. The manufacture of
such material at private plants necessitated its shipment to the filling
plant at Edgewood. The transportation of large quantities of highly
toxic gases seemed attended with great danger. The Director
General of Railroads ruled that all such shipments must be made by
special train, a very expensive method of transportation. Still more
serious objections were encountered in the attempt to enlist the co-
operation of existing firms. They recognized that the manufacture of
such material would be attended by very great danger; that the work
would be limited to the duration of the war; and that the processes
involved, as well as the plants necessary for carrying out their
processes, would have little post-war value. Moreover, such firms as
had the personnel and equipment were already over-worked. With a
few exceptions (notably the American Synthetic Color Company, the
Oldbury Electro-Chemical Co., Zinsser & Co., and the Dow Chemical
Company) they were unwilling to undertake work of this character on
any terms whatever.
Early in December, 1917, therefore, it was decided to erect, on
the site of the shell filling plant, such chemical plants as would be
necessary to furnish the toxic materials required for filling the shell.
The Arsenal is situated in an isolated district, twenty miles east of
Baltimore, Maryland, on the Pennsylvania Railroad, and comprises
3,400 acres. Since the main line of the Pennsylvania Railroad runs
on one side of the tract, while on another is the Bush River, only a
few miles from its mouth in Chesapeake Bay, the tract was ideally
situated for shipping. This site was referred to, at first, as
“Gunpowder Reservation,” but on May 4, 1918, the name was
officially changed to “Edgewood Arsenal.”
Fig. 6.—Edgewood Arsenal.

The upper view shows the site as it appeared
Oct. 24, 1917. The lower view shows the same as it
appeared nine months later.
Some idea of the extent of the work may be gained from the
following facts. On October 1, 1918, there were 233 officers, 6,948
enlisted men and 3,066 civilians engaged in work at Edgewood. 86
cantonments were built, accommodating about 8,500 men, while the
five officers’ barracks provided accommodations for 290. The
completed hospital unit consisted of 34 buildings, accommodating
420 patients under ordinary conditions. The total number of buildings
erected on the Arsenal grounds was 550. 14.8 miles of improved
roads were built, and 21 miles of standard gauge and 15 miles of
narrow gauge railway. A system furnishing 9.5 million gallons of salt
water and another furnishing two millions of fresh water daily were
successfully installed. Large power plants were built in connection
with the shell filling plants and the chlorine plant.
Plants for phosgene, chloropicrin, mustard gas, chlorine and
sulfur chloride were built and placed in successful operation. Most of
the raw materials, with the exception of sulfur chloride, were
obtained from commercial firms. The other gases and manufactured
materials used, such as phosphorus, tin and silicon tetrachlorides,
bromobenzylcyanide and arsenic derivatives were supplied by
various plants scattered through the East and Middle West States.
The raw materials used by the Arsenal in 1918 were as follows:
Salt 17,358,000 pounds
Bleach 42,384,000 “
Picric acid 3,718,000 “
Alcohol 3,718,000 “
Sulfur 24,912,000 “
Sulfur chloride 6,624,000 “
Bromine 238,000 “
Benzyl chloride 26,000 “
The production of toxic materials and the amount shipped
overseas in bulk follow:
Production, Shipped in
Bulk,
Pounds Pounds
Chlorine:
Liquid 5,446,000 2,976,000
Gaseous 2,208,000
Chloropicrin 5,552,000 3,806,000
Phosgene 3,233,070 840,000
Production, Shipped in
Bulk,
Pounds Pounds
Mustard gas 1,422,000 380,000
Bromobenzyl
10,000
cyanide
White phosphorus 2,012,000 342,000
Tin tetrachloride 2,012,000 212,000
Titanium
362,000
tetrachloride
For nearly a month previous to the signing of the Armistice, the
various plants at the Arsenal had shut down or were operated only to
an extent sufficient to maintain the machinery and equipment in good
working order, on account of the lack of shell into which to fill the
gas, so that the above figures do not at all represent maximum
productive capacity.
These plants will be described in the appropriate chapters.
The shell filling plant was really composed of several small
plants, each of which was made up of units radiating from a central
refrigeration plant which would serve all the units. Each unit could
then be fitted with machinery adapted for filling shell of a different
size, and for a particular gas. Moreover, an accident in one of the
units would in no way impair the working of the remainder.
The problem involved in the filling of a shell with toxic material
(which is always a liquid or a solid and never a gas under the
conditions in which it is loaded in the shell) is similar in a way to that
of filling bottles with carbonated water. In the development of plans
for the filling plant, many suggestions were obtained from a study of
the apparatus used in commercial bottling plants. It was necessary
to keep in mind not only the large number of shell to be filled, but
also the highly toxic character of the filling material to be used. It was
essential that the work of filling and closing the shell should be done
by machinery in so far as that was possible, and that the operation
should be carried out in a thoroughly ventilated room or tunnel,
arranged so that the machinery contained in the tunnel could be
operated from the outside. Special care was taken in closing the
shell, the closing being accomplished by motors actuated by
compressed air, which, in the closing process were driven until they
stalled. In this way a uniform closing torque was obtained. The final
results secured were admirable, as is evidenced by the fact, reported
by the Quartermaster Officer at Vincennes on November 15, 1918,
that not a single leaky shell had been found among the 200,000 shell
received up to that date.
Fig. 7.—A Typical Shell filling Plant at Edgewood

Arsenal.
Details of the filling process will be found in the chapter on
Phosgene.
Besides the ordinary gas filling plants (of which one was
completed and two were 80 per cent completed) there was a plant
for stannic chloride grenades, one for white phosphorus grenades,
and one for smoke shell also filled with phosphorus and a plant for
filling incendiary bombs.
Shell are designated by their diameter in inches or millimeters.
The approximate amount of toxic gas required for filling each type of
shell (10.5 per cent void) is as follows:
Phosgene, N. C.,[13] Mustard Gas,

Shell
Pounds Pounds Pounds
75 mm 1.32 1.75 1.35
4.7 inch 4.27 6.20 4.20
155 mm 11.00 15.40 10.35
8 inch 22.00 30.30 21.60
Livens 30.00
The gas grenades held 0.446 pound of stannic chloride, and the
smoke grenades held 0.67 pound of white phosphorus.
The only type of shell filled was the 75 mm. variety, because
either the shell of the other sizes or the accompanying boosters
(bursting charges) were not available.
The work done by the filling plant is shown by the following
figures, representing the number of shell, grenades, etc.
75 mm. Shell
Shipped
Filled
Overseas
Phosgene 2,009
N. C. 427,771 300,000
Shipped
Filled
Overseas
Mustard gas 155,025 150,000
Livens Drum
Phosgene 25,689 18,600
Grenades
White phosphorus 440,153 224,984
Tin tetrachloride 363,776 175,080
Incendiary Drop Bomb

Mark I. 542
Mark II. 2,104
The total monthly capacity of the filling plants at the date of the
Armistice was as follows:
Pounds
75 mm. shell 2,400,000
4.7 inch shell 450,000
155 mm. shell 540,000
6 inch shell 180,000
Gas grenade 750,000
Smoke grenade 480,000
Livens drum 30,000
One point relating to the casualties resulting from the work should
perhaps be mentioned here. The number of casualties should
change the mind of anyone who feels that men chose this work as
being “safe” instead of going to France. During the six months from
June to December there were 925 casualties, of which three were
fatal, two being due to phosgene and one to mustard gas. These
were divided among the different gases as follows:
Mustard gas 674
Stannic chloride 50
Phosgene 50
Chloropicrin 44
Chlorine 62
Other material 45
Of these 279 occurred during August, 197 during September and
293 during October. Since production stopped early in November,
there were only 14 during that month and three during December.
The Staff at Edgewood Arsenal at the signing of the Armistice
was as follows:
Commanding Officer Colonel Wm. H. Walker
Lt. Colonel George
Cahoon, Jr.
Administrative Officers
Lt. Col. Edward M. Ellicott
Lt. Col. Wm. C. Gallowhur
Lt. Col. Wm. McPherson
In Charge of Outside Major Adrian Nagelvoort
Plants Major Charles R. Wraith
Captain John D. Rue
Shell Filling Plant Lt. Col. Edwin M. Chance
Chlorine Plant Lt. Col. Charles Vaughn
Chemical Plants Major Dana J. Demorest
Chemical Laboratory Major William L. Evans
As the work of the Arsenal expanded it was necessary to
manufacture certain of the chemicals at outside plants. The men in
charge of these plants were:
Bound Brook, N. J. Lt. William R.
Chappell
Stamford, Conn. Lt. V. E. Fishburn
Hastings-on-Hudson, N. Major F. G. Zinnsser
Y.
Niagara Falls, N. Y. Major A. Nagelvoort
Buffalo, N. Y. Lt. A. W. Davison
Kingsport, Tenn. Lt. E. M. Hayden
Charleston, W. Va. Lt. M. R. Hoyt
Midland, Mich. Major M. G. Donk
Croyland, Pa. Capt. A. S. Hulburt
After the Armistice, Edgewood Arsenal was selected as the
logical home of the Chemical Warfare Service, and all the outside
activities of the Service were gradually closed up and the physical
property and files moved to Edgewood. At first the command of the
Arsenal was in the hands of Lt. Col. Fries, but when he was
appointed Chief of the Service, Major E. J. Atkisson, who had so
successfully commanded the First Gas Regiment, A. E. F., was
happily chosen his successor. At the present time (July 1, 1921), the
organization of Edgewood Arsenal is as follows:
Commanding Officer Major E. J. Atkisson
Executive Officer Major R. C. Ditto
Technical Director Dr. J. E. Mills
Chemical Division Mr. D. B. Bradner
Mechanical Division Mr. S. P. Johnson
Capt. E. G.
Plant Division
Thompson
Chemical Warfare School Major O. R. Meredith
Property Major A. M. Heritage
First Gas Regiment Major C. W. Mason

Conceptual Modeling 33rd International Conference ER 2014 Atlanta GA USA October 27 29 2014 Proceedings 1st Edition Eric Yu

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Conceptual Modeling 33rd International Conference ER 2014 Atlanta GA USA October 27 29 2014 Proceedings 1st Edition Eric Yu

Uploaded by

Copyright:

Available Formats

Conceptual Modeling 33rd International

Conference ER 2014 Atlanta GA USA

Intelligent Virtual Agents 14th International

Mobile Web Information Systems 11th International

Case Based Reasoning Research and Development 22nd

Knowledge Engineering and the Semantic Web 5th

System Analysis and Modeling Models and Reusability 8th

Computer Safety Reliability and Security 33rd

Simulation Modeling and Programming for Autonomous

HCI International 2014 Posters Extended Abstracts

ISSN 0302-9743 e-ISSN 1611-3349

Library of Congress Control Number: 2014950401

The 33rd edition of the International Conference on Conceptual Modeling (ER),

Finally, three highly interesting keynote presentations addressed important

July 2014 Eric Yu

Technical Program Co-chairs

Poster / Demo Chair

PhD Symposium Chair

Local Organization Chair

Steering Committee Liaison

Isabel F. Cruz University of Illinois, USA

Oscar Pastor Lopez University of Valencia, Spain

Bertossi, Leopoldo Cappiello, Cinzia

Horst, Andreas Navarro Perez, Antonio

Professor and Director

Abstract. Data Science is the study of the generalizable extraction

Antonio L. Furtado, Marco A. Casanova, and Simone D.J. Barbosa

Abstract. The work on Conceptual Modelling performed by our group

Keywords: Entity-Relationship Model, Information Systems, Planning,

Ontology and Conceptual Modeling Research Group (NEMO),

Abstract. This paper addresses the complexity of conceptual modeling

Keywords: Conceptual Modeling, Formal Ontology, Patterns,

Ontological Patterns, Anti-Patterns and Pattern Languages for

Data on the Web

On Designing Archiving Policies for Evolving RDF Datasets

Ontology-Based Spelling Suggestion for RDF Keyword Search . . . . . . . . . 57

Mapping Heterogeneous XML Document Collections to Relational

MKStream: An Eﬃcient Algorithm for Processing Multiple Keyword

Uncertain and Incomplete Data

TopCrowd – Eﬃcient Crowd-enabled Top-k Retrieval on Incomplete

Web Services Composition in the Presence of Uncertainty

Big Data, Graphs and Networks

Network Analytics ER Model – Towards a Conceptual View of Network

Model-Driven Design of Graph Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Privacy and Safety

From Conceptual Models to Safety Assurance . . . . . . . . . . . . . . . . . . . . . . . 195

Database Design for NoSQL Systems (Short paper) . . . . . . . . . . . . . . . . . . 223

Fixing Up Non-executable Operations in UML/OCL Conceptual

Generic Data Manipulation in a Mixed Global/Local Conceptual

New Modeling Languages and Applications

Nòmos 3: Legal Compliance of Roles and Requirements . . . . . . . . . . . . . . . 275

Towards an XBRL Ontology Extension for Management Accounting

Representing Hierarchical Relationships in INM (Short paper) . . . . . . . . . 297

Providing Foundation for User Feedback Concepts by Extending a

Towards a Conceptual Framework and Metamodel for Context-

Software Concepts and Strategies

Modelling Risks in Open Source Software Component Selection . . . . . . . . 335

Modelling and Applying OSS Adoption Strategies . . . . . . . . . . . . . . . . . . . . 349

Patterns and Narratives

Recall of Concepts and Relationships Learned by Conceptual Models:

Visual Maps for Data-Intensive Ecosystems (Short paper) . . . . . . . . . . . . . 385

Data Management for Enterprise Architecture

Modeling Organizational Alignment (Short paper) . . . . . . . . . . . . . . . . . . . 407