earticle

영어

Community detection is widely applied in many fields and k-clique community detection is one important detection method. There are many works on k-clique community detection. However, the work on analyzing the structure of k-clique community is rare. In this paper, we first give the definition of k-clique community tree and closed l-s-clique community, which could be used as the index of analyzing k-clique community. Then we give the definition of l-s-clique community pivot to describe the members playing the bridging roles in k-clique community. We analyze the properties of l-s-clique community and propose KCliqueTree algorithm based on the properties. This algorithm could efficiently generate k-clique community tree whose leaf nodes represent closed l-s-clique community. We also propose LSBridge algorithm to search l-s-clique community pivot. At last, we conduct case study on DBLP (Digital Bibliography & Library Project) dataset, which shows the availability of our definitions and algorithms.