Collaboration through deliberative dialogues

Abstract

In a multi-agent system, a group of individ- uals interact in a social context in order to boost their capabilities and enhance global per- formance. Each individual's action repertoire may be reduced, but it's social capability allows it to interact with other agents and obtain collab- oration. This work offers an alternative for knowledge representation in a system of collaborative BDI agents and presents an interaction protocol based on dialogues. The capacity to interact affects the behavioral model of a BDI agent that must consider the possibility of offering and soliciting collaboration. Thus, we propose an algorithm that models the behavior of a collaborative BDI agent.