Collaboration among Agents in a Virtual Collaborative Learning Environment

摘要

Among the learning effects which are expected by collaborative learning, in this study we concentrate on problem-solving knowledge fixing via 'Learning by Teaching' and problem- solving knowledge asquisition via 'Learning by Observing'. We propose a virtual collaborative learning environment called CASLE (Companion AgentS Learning Environment), embedding Companion Agents which have each of these learning effects as Teaching Strategies. These Companion Agents' behavior is based on a 'Student Model', which represents the understanding states of a Real Student, and 'Learning Rules', which represent what level of knowledge a real student aquires by experiencing situations in the collaborative learning environment, according to each Teaching Strategy A Facilitator Agent manages these Models and controls the Companion Agents' behavior. Moreover, it identifies faults of the learning rules used by the real student from the Learning Rules, and it points them out to him. In order to realize such a learning environment, in this study we model each agent and the collaboration between agents. In this paper, we describe these models in detail.