Simulating rational goal-directed behaviour using a logic-based programming language for multi-agent systems

Abstract

An essential aspect of autonomous agents is that they must display proactive behaviour. Designing such software then requires explicit consideration of the goals the agent ought to achieve, and similarly its implementation also needs to be based on explicit representations of such goals. This is part of the reason why the BDI (Belief-Desire- Intention) agent architecture [16, 17] has since the early 90's been the best known architecture for developing software agents. As the BDI notions are also used in "folk psychology" (i.e., how people ordinarily refer to other people's behaviour) it also makes it useful for modelling goal-directed human behaviour [12].