Community Detection by Affinity Propagation

Abstract

Community structure in networks indicates groups of ver- tices within which are dense connections and between which are sparse connections. Community detection, an important topic in data mining and social network analysis, has attracted considerable research interests in recent years. Motivated by the idea that community detection is in fact a clustering problem on graphs, we propose several similarity met- rics of vertex to transform a community detection problem into a cluster- ing problem, and further adopt a recently-proposed clustering method, namely Affinity Propagation, to extract communities from graphs. We demonstrate that the method achieves significant quality in detecting community structures in both computer-generated and real-world net- work data in near-linear time. Furthermore, the method could automat- ically determine the number of communities.