Naïve Bayesian Learning based Multi Agent Architecture for Telemedicine

Agent-based systems are one of the most vibrant and important areas of the research and development to have emerged in Information Technology in recent years. They are one of the most promising approaches for designing and implementing autonomous, intelligent and social software assistants capable of supporting human decision-making. These kinds of systems are believed to be appropriate in many aspects of the healthcare domain. As a result, there is a growing interest of researchers in the application of agent-based techniques to problems in the healthcare domain. The adoption of agent technologies and multi-agent constitutes an emerging area in bioinformatics. Multi-agent based medical diagnosis systems may improve traditionally developed medical computational systems and may also support medical staff in decision-making. In this paper, we simulate the multi agent system for cancer classification. The proposed architecture consists of service provider agents as upper layer agent, coordinator agent as middle layer agent and initial agent lowest layer agent. Coordinator agent serves as matchmaker agent that uses Naïve Bayesian learning method for obtaining general knowledge and selects the best service provider agent using matchmaking mechanism. Therefore this system can reduce the communication overhead between agents for sending messages and transferring data and can avoid sending the problem to irrelevant agents.