Emergence and stability of collaborations among rational agents

Abstract

Autonomous agents interacting in an open world can be considered to be primarily driven by self interests. In this paper, we evaluate the hypotheses that self-interested agents with complementary expertise can learn to recognize cooperation possibilities and develop stable, mutually beneficial coalitions that is resistant to exploitation by malevolent agents. Previous work in this area has prescribed a strategy of reciprocal behavior for promoting and sustaining cooperation among self-interested agents. That work have considered only the task completion time as the cost metric. To represent more realistic domains, we expand the cost metric by using both the time of delivery and quality of work. In contrast to the previous work, we use heterogeneous agents with varying expertise for different job types. This necessitates the incorporation of the novel aspect of learning about other's capabilities within the reciprocity framework. We also present a new mechanism where agents base their decisions both on historical data as well as on future interaction expectations. A decision mechanism is presented that compares current helping cost with expected future savings from interaction with the agent requesting help.