Developing a multi-agent model for distributed knowledge systems

Development of knowledge sharing, problem solving, and communication facilities is a key issue for the next generation knowledge systems for complex applications in distributed cooperative environments. The authors examine requirements of a multi-agent model for such distributed systems and propose the concept of distributed explanation-based learning (DEBL). DEBL is based on the idea of explanation based learning (EBL) extended into distributed environments. The authors also demonstrate the features of DEBL by a small example. This example suggests that DEBL can be a simple and extensible model for the learning mechanisms to improve problem solving and communication capabilities of multi-agent in distributed environments