A Multi-agent and Data Mining Model for TCM Cases Knowledge Discovery

In the past twenty years, agents and data mining have emerged separately as two prominent, dynamic and exciting research areas. In recent years, an increasingly remarkable trend in both areas is that integrating agents into data mining systems in order to extract knowledge from the distributed/ heterogeneity of data. We propose a TCMMADM which is a multi-agent and data mining model to extract knowledge from database on TCM cases, which the agents use different methods to handle different types or parts of information in heterogeneous data sets. The results are combined to get integrated results In this model We propose three types of agents, firstly is manager agent which create, suspect, resume miner agent and cooperator agent, secondly is miner agent, which apply multiple algorithms to data sets for extract knowledge, finally is cooperator agent, which is to integrated partial and local knowledge to more high level knowledge.