Michael Rovatsos
ABSTRACT
This paper introduces InFFrA, a novel method for the analysis and design of multiagent systems that is based on the notions of interaction frames and framing. We lay out a conceptual framework for viewing multiagent systems (MAS) as societies consisting of socially intelligent agents that record and organise their interaction experience so as to use it strategically in future interactions. We also provide criteria for the class of MAS InFFrA is suited for. The benefits of our approach are that it helps to understand and develop socially intelligent agents as well as to identify shortcomings of existing MAS. The method is evaluated through the analysis of an opponent classification heuristic that is used to optimise strategic behaviour in multiagent games, and interesting issues for future research are discussed.
- W. Brauer, M. Nickles, M. Rovatsos, G. Weiß, and K. F. Lorentzen. Expectation-Oriented Analysis and Design. In Proc. AOSE'2001, volume 2222 of LNCS. Springer-Verlag, Berlin, 2001 Google Scholar
Digital Library
- D. Carmel and S. Markovitch. Learning and using opponent models in adversary search. Technical Report 9609, Technion, 1996Google Scholar
- D. Carmel and S. Markovitch. Learning models of intelligent agents. In Proc. AAAI'1996, pages 62--67, Menlo Park, CA, 1996. AAAI Press/MIT Press Google ScholarDigital Library
- D. Fudenberg and J. Tirole. Game Theory. The MIT Press, Cambridge, MA, 1991Google Scholar
- E. Goffman. Frame Analysis: An Essay on the Organisation of Experience. Harper and Row, New York, NY, 1974Google Scholar
- J. L. Kolodner. Case-Based Reasoning. Morgan Kaufmann, San Francisco, 1993 Google ScholarDigital Library
- R. D. Luce and H. Raiffa. Games and Decisions. John Wiley & Sons, New York, NY, 1957Google Scholar
- M. Rovatsos. Interaction frames for artificial agents. Technical Report Research Report FKI-244-01, Technical University of Munich, 2001Google Scholar
- M. Rovatsos and M. Wolf. Towards Social Complexity Reduction in Multiagent Learning: the AdHoc Approach. Technical Report SS-02-02, AAAI Press, Stanford, CA, 2002Google Scholar
- C. Watkins and P. Dayan. Q-learning. Machine Learning, 8:279--292, 1992 Google ScholarDigital Library
Index Terms
- An approach to the analysis and design of multiagent systems based on interaction frames
Recommendations
Organizing Multiagent Systems
Despite all the research done in the last years on the development of methodologies for designing MAS, there is no methodology suitable for the specification and design of MAS in complex domains where both the agent view and the organizational view can ...
Multiagent Reinforcement Social Learning toward Coordination in Cooperative Multiagent Systems
Most previous works on coordination in cooperative multiagent systems study the problem of how two (or more) players can coordinate on Pareto-optimal Nash equilibrium(s) through fixed and repeated interactions in the context of cooperative games. ...
Comments