|
GDM: Group Decision Making in Multi-Agent Systems |
|
Organization responsible: Thales Research & Technology France People involved: Project description: This project aims to research techniques and models for collaborative decision making situations in multi-actor environments, with a specific focus on coalition formation, consensus formation and negotiation. The results can be used for designing collaborative decision making strategies for synthetic actors.
In complex organizations, many agents need to collaborate in decision making processes, and they may have wildly different beliefs on the best decision to take. Especially when more than one team or organization is involved in the process, decision making agent usually have different and conflicting goals and preferences, and a consensus amongst participating agents must somehow be reached. Moreover, these types of consensus making suffer from bounded rationality due to the lack of knowledge (partial situation awareness) and possible contradictory expert opinions. This project aims to research and model these types of decision making, in order to design multi-agent communities that are capable of, or can offer support in group decision making.
This project aims to research the capability of a set of agents (especially human agents) in making collective decisions. The proposed agents are cooperative beings, capable of having private objectives. The agents are endowed with representations of the world, mechanisms of deduction, decision making and interaction means. Collaborative decision making will be studied in an incremental fashion approach where we define more or less independent agents and the negotiation protocols which allow them to build one or several solutions which satisfy their utility functions, their preferences, their private objectives, etc. This project is focused around two areas of group decision making: coalition formation and negotiation strategies. Publications:
|
