Research on the Key Techniques of Heterogeneous Agents Supported and Personalized Teaching and Resource Integration

The traditional Agent teaching and resource integration encounters troubles for it is only suited to closed teaching systems (systems do not permit Agents to dynamically join), but not suited to the heterogeneous Agents (Agents with different standards) participated, open (permit Agents to dynamically join), non -deterministic (with self-interest Agents), intelligent (Agents with mental states), and personalized (according to personal requirements) teaching environment. In this paper, aiming at the virtual community/VO based policy-driven teaching as a main line, and with the purpose to construct an open, heterogeneous Agents supported and cooperative teaching infrastructure, and the personalized teaching and resource integration application, we propose the key techniques including teaching ontology modeling, teaching mediation, teaching negotiation, and the teaching law and government. We design the experiment using the prototype of intelligent teaching and learning system based on two parts platform, and using the success rate of the task performance to compare the proposed system with the other system. The experiment shows that the proposed system has an average success rate enhancement of 60.9%, and supports heterogeneous Agents participated, open, intelligent, and personalized teaching and resource integration very well.