Agent Reasoning Based On Trust And Reputation

In the area of intelligent systems development some deterministic or nondeterministic decision algorithms and mechanisms should be used to enable agents to behave intelligently. We are trying to enhance agent reasoning and especially agent decision making with a usage of trust and reputation of particular intelligent elements (agents) as well as some social groups. There can be large agent societies, where collaboration between agents is the best way and sometime the only possibility to achieve non-trivial goals. Often it is very difficult to find best counterparts for collaboration. Our approach works with trust and reputation principles which are inspired from realworld societies and we try to shift them into artificial societies to make their interaction and cooperation more effective.