1 / 9

Decentralized, Agent Based, and Social Approaches to User Modeling (DASUM)

Decentralized, Agent Based, and Social Approaches to User Modeling (DASUM). Peter Dolog and Julita Vassileva http://www.l3s.de/~dolog/dasum/. DASUM. Decentralized: user modelling information is distributed everywhere, in different formats/representations

gwyn
Download Presentation

Decentralized, Agent Based, and Social Approaches to User Modeling (DASUM)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Decentralized, Agent Based, and Social Approaches to User Modeling (DASUM) Peter Dolog and Julita Vassileva http://www.l3s.de/~dolog/dasum/

  2. DASUM • Decentralized: user modelling information is distributed everywhere, in different formats/representations • Agent-Based: agents/applications collect and process on demand, for particular purpose, in particular context, under resource constraints • Social: these agents are involved in social relationships, groups, teams that impact the purpose and context P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  3. Statistics • 19 submissions • 6 full papers accepted (31,5%) • 6 short papers accepted (one withdrawn) • Organization of workshop: • Morning paper presentations • Afternoon posters and discussions P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  4. Alfred Kobsa, University of California at Irvine, USA Antonio Krueger, DFKI, Germany Daniel Kudenko, University of York, UK Gord McCalla, University of Saskatchewan, Canada Daniel Olmedilla, L3S Research Center, University of Hannover, Germany Olayide Olorunleke, University of Saskatchewan, Canada Fiorella de Rosis, University of Bari, Italy Boris De Ruyter, Philips Research, The Netherlands Michael Sintek, DFKI, Germany Amy Soller, Institute for Defense Analyses, USA Thomas Tran, University of Ottawa, Canada Marianne Winslett, University of Illinois at Urbana-Champaign, USA PC Members: Liliana Ardissono, University of Torino, Italy Lora Aroyo,Eindhoven University of Technology, The Netherlands Mathias Bauer, DFKI, Germany Peter Brusilovsky, University of Pittsburgh, PA, USA Susan Bull, University of Birmingham, UK Keith Cheverst, University of Lancaster, UK Robin Cohen, University of Waterloo, Canada Nadia de Carolis, University of Bari, Italy Ludger van Elst, DFKI, Germany Elena Gaudioso,University for Distance Learning, Spain Piotr Gmytrasiewicz University of Illinois, Chicago, USA Dominik Heckmann, DFKI, Germany Judy Kay, University of Sydney, Australia Organization P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  5. Topics Covered by Papers • Assembling user model from distributed fragments • User identification in decentralized and mobile environments • Architectures for decentralized user modeling • Recommendations in distributed environments • Clustering users into groups in distributed environments • Social aspects in group models P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  6. Issues Discussed • Physical vs. conceptual distribution • How to deal with contradictions • Who, when, where and how will decide? • Resolving contradictions or information integration/reconcilation? • How to reconcile user model? • Are ontologies and semantic web the silver bullet? • How to map between heterogeneous representations? • Precomputing user model data or computing it on the fly (on demand, in context)? • Protocols for information exchange: how they should be designed? P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  7. ...More Issues • Ensuring common understanding: • Centralized ontology versus local meaning negotiation • Approaches: Negotiation vs. Machine Learning/data mining • Emphasis on computation (in context for given purpose) rather than on storing um data; reuse of user model computing units • Will social/sociological research help us to proceed? • Privacy, trust, scrutability • How problems relate to other disciplines – distributed databases, enterprise information systems, semantic web... P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  8. How this changes the UM agenda? • Emphasis on: • Data integration • In context, not universal • Procedures for modelling, not representation of models • Incremental semantic agreement among agents • Computational tradeoffs • On demand versus pre-computing • What to store P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

  9. Further Steps • Follow up workshop(-s) in future • Edited collection • Please, all participants and people interested to receive information, send an email to : dolog@l3s.de or jiv@cs.usask.ca P. Dolog and J.Vassileva, DASUM Workshop Summary (w2) UM'05, August 2005

More Related