Lecture Notes in Artificial Intelligence 2253 SubseriesofLectureNotesinComputerScience EditedbyJ.G.CarbonellandJ.Siekmann Lecture Notes in Computer Science EditedbyG.Goos,J.Hartmanis,andJ.vanLeeuwen 3 Berlin Heidelberg NewYork Barcelona HongKong London Milan Paris Tokyo Takao Terano Toyoaki Nishida Akira Namatame Syusaku Tsumoto Yukio Ohsawa Takashi Washio (Eds.) New Frontiers in Artificial Intelligence Joint JSAI 2001 Workshop Post-Proceedings 1 3 SeriesEditors JaimeG.Carbonell,CarnegieMellonUniversity,Pittsburgh,PA,USA Jo¨rgSiekmann,UniversityofSaarland,Saarbru¨cken,Germany VolumeEditors TakaoTerano YukioOhsawa TheUniversityofTsukuba,GraduateSchoolofBusinessScience E-mail:{terano/osawa}@gssm.otsuka.tsukuba.ac.jp ToyoakiNishida TheUniversityofTokyo,SchoolofEngineering E-mail:[email protected] AkiraNamatame NationalDefenseAcademy,Dept.ofComputerScience E-mail:[email protected] SyusakuTsumoto ShimaneMedicalUniversity,SchoolofMedicines E-mail:[email protected] TakashiWashio OsakaUniversity,TheInstituteofScientificandIndustrialResearch E-mail:[email protected] Cataloging-in-PublicationDataappliedfor DieDeutscheBibliothek-CIP-Einheitsaufnahme Newfrontiersinartificialintelligence:jointJSAI2001workshoppost proceedings/TakaoTerano...(ed.).-Berlin;Heidelberg;NewYork; Barcelona;HongKong;London;Milan;Paris;Tokyo:Springer,2001 (Lecturenotesincomputerscience;Vol.2253:Lecturenotesin artificialintelligence) ISBN3-540-43070-9 CRSubjectClassification(1998):I.2,H.2.8,H.3,F.1,H.4,H.5.2,I.5,J.1,J.3,K.4.3-4 ISSN0302-9743 ISBN3-540-43070-9Springer-VerlagBerlinHeidelbergNewYork Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting, reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965, initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer-Verlag.Violationsare liableforprosecutionundertheGermanCopyrightLaw. Springer-VerlagBerlinHeidelbergNewYork amemberofBertelsmannSpringerScience+BusinessMediaGmbH http://www.springer.de ©Springer-VerlagBerlinHeidelberg2001 PrintedinGermany Typesetting:Camera-readybyauthor,dataconversionbyPTP-Berlin,StefanSossna Printedonacid-freepaper SPIN:10845999 06/3142 543210 Preface TheJapaneseSocietyforArtificialIntelligence(JSAI)wasestablishedinJuly 1986. Since then, we have held conferences every year. Although JSAI is the second largest community in the world focusing on the area of Artificial In- telligence and we have over 3,000 members, the importance of the research presented and discussions held at the annual conferences has not been fully recognized in the Artificial Intelligence communities elsewhere in the world, partly because most presentations are made in the Japanese language. The- refore, the program committee of the Fifteenth Annual Conference of JSAI decided to open the door to the world and hold international workshops du- ring the conference on May 20th and 25th, 2001 in Matsue City, Japan. The workshop proposals were gathered from the members of JSAI. We accepted the following up-to-date and exciting topics: 1) Social Intelligence DesignchairedbyProf.ToyoakiNishida,UniversityofTokyo,2)Agent-Based ApproachesinEconomicandSocialComplexSystemschairedbyProf.Akira Namatame,NationalAcademyofDefense,3)RoughSetTheoryandGranular Computing chaired by Prof. Shusaku Tsumoto, Shimane Medical University, 4) Chance Discovery chaired by Prof. Yukio Osawa, and 5) Challenge in Knowledge Discovery and Data Mining chaired by Prof. Takashi Washio, Osaka University. These workshops were highly welcome and successful. A total of 116 people in Japan and 30 researchers from abroad participated in them. This volume of the proceedings contains selected papers presented at the workshops. The contents of the volume are divided into five parts, each of which corresponds to the topics of the workshops. Each paper was strictly reviewedbythecommitteemembersoftheworkshops.Theyalsocoverrecent divergent areas of artificial intelligence. We believe that the volume is highly useful for both researchers and practitioners who have interests in recent advances in artificial intelligence. October 2001 Takao Terano JSAI Workshops as International Trends Looking at the current economic, political, and ecological situations, we be- come aware of the dynamic environment surrounding all human activities. Hand in hand, the expansion of the World Wide Web is activating the whole globe as an information system including humans, computers, and networks. TheworkshoptopicsassociatedwithJSAI2001weredesignedtohitsuch world wide trends. Social Information Designs are needed to aid the mu- tual progress of human society and various kinds of information flows. The Agent-Based Simulations consider social behavior from the aspect of eco- nomics, with the up-to-date viewpoint of complexity. Rough Set Theories mayachieveabreakthroughwithregardtodealingwithuncertainrealworld events on the basis of established theories. Chance Discovery is a new direc- tionproposedbyJapaneseresearches,forhelpingpeopleandagentsbeaware of novel information, significant for their own decisions in dynamic environ- ments. KDD-Challengers are responding to requirements for new knowledge to be obtained from new data in new social situations. I am sure the selected papers from these first international workshops associated with JSAI will win the attention of people from several different areas of research, not only artificial intelligence but also social sciences and other areas looking into the future of human life. A piece of good news for thosereadersisthatJSAIisbecomingincreasinglyinternational,aftermany yearsasasemi-domesticJapaneseAIcommunity.Withthefoundationoffive workshop themes this year, the new generation of AI researchers is finding new problems and new solutions in the creative atmosphere. On behalf of all the workshop organizers, I wish to draw readers’ attention to forthcoming international JSAI events. Before beginning the contents, let us express our gratitude to the great support given by the co-editors who organized each workshop, all authors and audiences, JSAI committee members, Shimane prefecture and Matsue city,andJun’ichiroMorioftheUniversityofTokyowhoseoperationsgreatly aided this publication. October 2001 Yukio Ohsawa Table of Contents Part I. Social Intelligence Design 1. Social Intelligence Design – An Overview Toyoaki Nishida ............................................. 3 1.1 Introduction ........................................... 3 1.2 Horizon of Social Intelligence Design ...................... 4 1.2.1 Methods of Establishing the Social Context.......... 6 1.2.2 EmbodiedConversationalAgentsandSocialIntelligence 6 1.2.3 Collaboration Design ............................. 7 1.2.4 Public Discourse ................................. 8 1.2.5 Theoretical Aspects of Social Intelligence Design ..... 8 1.2.6 Evaluations of Social Intelligence ................... 9 1.3 Concluding Remarks ................................... 10 2. FaintPop: In Touch with the Social Relationships Takeshi Ohguro, Kazuhiro Kuwabara, Tatsuo Owada, and Yoshinari Shirai ............................................. 11 2.1 Social Intelligence Design for Communications ............. 11 2.2 In Touch with the Social Relationships .................... 13 2.3 Initial Experiment...................................... 16 2.4 Conclusion and Related Works ........................... 17 3. From Virtual Environment to Virtual Community A. Nijholt................................................... 19 3.1 Introduction ........................................... 19 3.2 Towards Multi-user Virtual Worlds ....................... 19 3.2.1 Interacting Embodied Personalities ................. 20 3.2.2 Embodied Personalities in Virtual Worlds ........... 21 3.3 Building a Theater Environment ......................... 23 3.4 Interacting about Performances and Environment .......... 24 3.5 Towards a Theater Community........................... 25 X Table of Contents 4. Collaborative Innovation Tools John C. Thomas ............................................. 27 4.1 Importance of Collaboration: Practical and Scientific........ 27 4.2 New Technological Possibilities........................... 29 4.3 Work of the Knowledge Socialization Group ............... 31 5. Bricks & Bits & Interaction R. Fruchter ................................................. 35 5.1 Introduction ........................................... 35 5.2 Visibility,Awareness,andInteractioninVideoconferenceSpace 36 5.3 Mobile Learners in E-learning Spaces ..................... 39 5.4 Emerging Changes Influenced by Bricks & Bits & Interaction 41 6. A Distributed Multi-agent System for the Self-Evaluation of Dialogs Alain Cardon................................................ 43 6.1 Introduction ........................................... 43 6.2 System General Architecture............................. 44 6.3 Representation of the Semantic of the Communication Act... 45 6.4 Semantic Traits and Agents.............................. 46 6.5 Aspectual Agent Organization ........................... 46 6.6 The Emerging Meaning of the Communication: The Morphological Agent Organization.................... 48 6.7 Interpretation of the Morphological Organization: The Evocation Agents .................................. 49 6.8 Conclusion ............................................ 50 7. Public Opinion Channel: A System for Augmenting Social Intelligence of a Community Tomohiro Fukuhara, Toyoaki Nishida, and Shunsuke Uemura ..... 51 7.1 Introduction ........................................... 51 7.2 Communication Costs................................... 52 7.3 POC Prototype System ................................. 53 7.3.1 POC Server ..................................... 53 7.3.2 POC Client: POCViewer .......................... 54 7.4 Evaluation............................................. 57 7.5 Discussion ............................................. 57 7.5.1 Automatic Broadcasting System.................... 57 7.5.2 POC and Narrative Intelligence .................... 58 7.6 Conclusion ............................................ 58 Table of Contents XI 8. Enabling Public Discourse Keiichi Nakata............................................... 59 8.1 Introduction ........................................... 59 8.2 Enabling Individuals to Collect and Exchange Information and Opinions .......................................... 60 8.3 Raising Social Awareness through Position-Oriented Discussions ............................................ 62 8.3.1 Positioning-Oriented Discussion Interface............ 63 8.4 Towards “Social Intelligence Design”...................... 64 8.5 Concluding Remark..................................... 65 9. Internet, Discourses, and Democracy R. Luehrs, T. Malsch, and K. Voss ............................. 67 9.1 Introduction ........................................... 67 9.2 Online Support for Democratic Processes.................. 67 9.3 A Novel Participation Methodology....................... 69 9.4 System Design ......................................... 72 10. How to Evaluate Social Intelligence Design Nobuhiko Fujihara .......................................... 75 10.1 Computer Networked Community as Social Intelligence ..... 75 10.2 The Importance of Control Condition in Evaluating Social Intelligence Design...................................... 76 10.3 How to Evaluate POC .................................. 77 10.4 Future Works .......................................... 81 Part II. Agent-Based Approaches in Economic and Social Complex Systems 11. Overview Akira Namatame ........................................... 85 12. Analyzing Norm Emergence in Communal Sharing via Agent-Based Simulation Setsuya Kurahashi and Takao Terano ......................... 88 12.1 Introduction ........................................... 88 12.2 Related Work on Studies of Norms ....................... 89 12.3 Artificial Society Model TRURL .......................... 90 12.3.1 Agent Architecture ............................... 90 12.3.2 Communication and Action Energy................. 91 12.3.3 Inverse Simulation................................ 91 12.4 Experiments ........................................... 92 XII Table of Contents 12.4.1 An Amount of Information in Each Society .......... 92 12.4.2 Emergence and Collapse of a Norm ................. 93 12.4.3 Emergence and Control of Free Riders .............. 94 12.4.4 Information Gap ................................. 95 12.4.5 Discussion....................................... 96 12.5 Conclusion ............................................ 97 13. Toward Cumulative Progress in Agent-Based Simulation Keiki Takadama and Katsunori Shimohara .................... 99 13.1 Introduction ........................................... 99 13.2 Can We Assist Cumulative Progress? ..................... 100 13.2.1 Problems in Agent-Based Approaches............... 100 13.2.2 Points for Cumulative Progress..................... 100 13.2.3 Cumulative Progress in Current Projects ............ 101 13.3 Exploring Key Elements................................. 101 13.3.1 Interpretation by Implementation .................. 102 13.3.2 Applications of IbI Approach ...................... 103 13.4 Discussion ............................................. 104 13.4.1 Cumulative Progress.............................. 104 13.4.2 Potential of Our Approach ........................ 105 13.5 Conclusions............................................ 107 14. Complexity of Agents and Complexity of Markets Kiyoshi Izumi .............................................. 110 14.1 Introduction ........................................... 110 14.2 The Efficient Market Hypothesis Seen from Complexity ..... 111 14.3 Artificial Market Model ................................. 112 14.3.1 Expectation ..................................... 112 14.3.2 Order........................................... 113 14.3.3 Price Determination .............................. 113 14.3.4 Learning ........................................ 113 14.4 Simulation Result ...................................... 114 14.4.1 Merit of Complicating a Prediction Formula ......... 114 14.4.2 The Demerit in the Whole Market.................. 115 14.4.3 Development of the Complexity of a Market ......... 115 14.5 New Efficient Market Hypothesis ......................... 118 14.6 Conclusion ............................................ 119