Core Nodes Detection Based on Frequent Itemsets of Graph

This paper concentrates on the detection of core nodes in the crime network,but as a basis,it models the network as a graph and presents the algorithm of GBMFS(graph based maximum frequent set),which combines the mining of association rules with bidirectional search strategy and can be used to discover the most frequent itemsets in a graph.After getting several snaps of social network in different time and integrating the discovery of quasi-clique with GBMFS,the algorithm of discovering core nodes in these snaps is achieved.At last the algorithm is applied in a real system.