History Based Coalition Formation in Hedonic Context Using Trust
This work addresses coalition formation for multi-agent systems, but it appears incremental as it builds on previous models by adding trust and risk behaviors.
The paper tackles the problem of coalition formation in hedonic contexts by allowing agents to leave coalitions with restrictions, using trust values to compute expected utility and introducing risk behaviors for coalition initiators, with results analyzed through simulations.
In this paper we address the problem of coalition formation in hedonic context. Our modelling tries to be as realistic as possible. In previous models, once an agent joins a coalition it would not be able to leave the coalition and join the new one; in this research we made it possible to leave a coalition but put some restrictions to control the behavior of agents. Leaving or staying of an agent in a coalition will affect on the trust of the other agents included in this coalition. Agents will use the trust values in computing the expected utility of coalitions. Three different risk behaviors are introduced for agents that want to initiate a coalition. Using these risk behaviors, some simulations are made and results are analyzed.