摘要：In order to overcome the difficulty of cognitive and mining for large-scale multi-relationship system, we take the advantages of intelligent computing of FCM (Fuzzy Cognitive Map) and update the traditional FCM to MRFCM (Multi-Relationship based Fuzzy Cognitive Map) for multi-relationship data mining. The state mining of hyper-node and multi-relationship inference of MRFCM are studied and implemented. The steady of the mining is proved and the experimental results of multi-relationship classification demonstrate the efficiency of MRFCM.
