Clustering dense graphs: A web site graph paradigm

Abstract

Typically graph-clustering approaches assume that a cluster is a vertex subset such that for all of its vertices, the number of links connecting a vertex to its cluster is higher than the number of links connecting the vertex to the remaining graph. We consider a cluster such that for all of its vertices, the number of links connecting a vertex to its cluster is higher than the number of links connecting the vertex to any other cluster. Based on this fundamental view, we propose a graph-clustering algorithm that identifies clusters even if they contain vertices more strongly connected outside than inside their cluster; hence, the proposed algorithm is proved exceptionally efficient in clustering densely interconnected graphs. Extensive experimentation with artificial and real datasets shows that our approach outperforms earlier alternate clustering techniques.