Introspection as Control in Result-Sharing Assumption-Based Reasoning Agents

Introspection refers to an agent’s knowledge of its own beliefs. In this paper we explore the use of introspection as a mechanism for control in result-sharing cooperative assumptionbased reasoning systems. Contrary to the current trend towards building agents with mechanisms for selective communication, our experiments indicate there exists an interesting class of cooperative distributed problem solving systems in which agents may broadcast results without regard to selection. Recipient agents avoid penalties typically associated with such broadcast strategies using introspection and a history of knowledge state to focus problem solving.