A Novel Clustering Algorithm for Graphs

Graph or network clustering is one of the fundamental multimodal combinatorial problems that have many applications in computer science. Many algorithms have been devised to obtain a reasonable approximate solution for the problem. Current approaches, however, suffer from the local optimum drawback and then have difficulty splitting two clusters with very confused structures. In this paper we propose a novel genetic-based algorithm incorporating with Modularity(Q(N)) fur the quality of partitioning of graphs. The theoretical analysis and experimental results on synthetic and real networks demonstrate superior performance over Newman's fast agglomerative algorithms in accuracy.