A Cross Iterative Algorithm for Mining Communities in Bipartite Network

Nowadays, it is of great significance to analyze the community structure for analyzing bipartite networks. Through the unremitting efforts of people, now many different community mining methods have arisen. But most of the methods need to ensure the number of the communities before mining, and relationship between the final community is one to one. This paper presents a cross iterative algorithm for mining communities in bipartite network. The first step is to determine the core of the community by looking complete sub-graph of the bipartite graph, after determining the core, alternately iterated the algorithm according to the core until the results is convergence. The last cores after the iteration are the communities of the bipartite network. Experimental results show that the algorithm runs excellent in a bipartite graph, and can mine many to many communities with higher quality.