Identifying community structures in networks with seed expansion

Abstract

Real-world networks naturally contain a lot of communities. Identifying the community structures is a crucial endeavor to analyze the networks. Here, we propose a novel algorithm which finds the community structures from seed expansion. Its expansion process bases on the transmissive probabilities coming from seed vertices and the modularity Q function which is firstly defined by Newman et al. The experimental evaluation is conducted on real-world networks. The evaluation shows that our algorithm has good results in quality. © Springer-Verlag Berlin Heidelberg 2010.