Multi-Agent Architecture for Intelligent Tutoring Systems

This paper shows a multi-agent architecture based on reactive agents for an intelligent tutoring system (ITS). The global system behavior is modeled taking into account the mental models and the cognitive task analysis. We present the basic characteristics of the reactive system in terms of reactive robotics, where they started. Next we introduce some definitions and schemes in order to characterize the multi-agent architecture. Finally, we present an ITS example called Makatsina Makatsina means tutor in TOTONACA, a Mexican pre-Spanish language. which teaches the skills necessary to solve a truss analysis problem by the joint method. This domain is an integration skill.