An agent belief system for multi-agent systems

Within the field of DAI, an agent must cooperate with other agents, the research on agent modelling is important. In this paper, a multi-agent system, MAPE (Multi-agent Processing Environment), is described, and the belief state model for an autonomous agent in the MAPE is presented. Our main focus is on the agent´s structure, agent´s belief and the connection between communication among agents and the changes of their belief states in the multi-agent societies where knowledge communication plays a key role among agents. We also propose some domain independent defaults for cooperating problem-solving in the communication situation by passing messages