One of the basic issues to face in Multi-Agent Systems is effectively implementingcooperation among agents. To this aim, a not trivial problem has to be solved:Among a (possible large) number of agents, how to detect which agents are promising candidates for cooperation? In this chapter a new approach for dealing with this problem is proposed. We define a formal model for representing agents and a number of semantic properties exploited for detecting fruitful cooperation. On the basis of this model we design a Multi-Agent System, called SPY, capable of managing andsupporting cooperation in the agent community. The system learns semantic propertiesby monitoring the user behavior in such a way that it adapts its response to userexpectation.
Finding the Best Agent for Cooperation
BUCCAFURRI F;ROSACI D;SARNE' G
2004-01-01
Abstract
One of the basic issues to face in Multi-Agent Systems is effectively implementingcooperation among agents. To this aim, a not trivial problem has to be solved:Among a (possible large) number of agents, how to detect which agents are promising candidates for cooperation? In this chapter a new approach for dealing with this problem is proposed. We define a formal model for representing agents and a number of semantic properties exploited for detecting fruitful cooperation. On the basis of this model we design a Multi-Agent System, called SPY, capable of managing andsupporting cooperation in the agent community. The system learns semantic propertiesby monitoring the user behavior in such a way that it adapts its response to userexpectation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
