Cheng-Gang Bian, Wen Cao, Gunnar Hartvigsen
We are investigating the cooperative behavior of intelligent agents in a large group. A cooperative intelligent agent is a user’s personal consultant that can take care of problems the user got, and solve them either by itself or via cooperation. We have proposed a twin-base (cooperator-base u task-base) agent modeling technique for individual agents to monitor activities of the others in order to achieve efficient cooperation in a small agent group. While the group grows in size and complexity, an activity-oriented hierarchical organization structure, club-community-society, has been studied. In this organization, each agent shows its activities to the outside world, and these activities are symbolic categories for the twin-base guided heuristic searches in peer locating processes.