Semidefinite programming based community detection for node-attributed networks and multiplex networks

Abstract

Community detection is an effective exploration technique for analyzing networks. Most of the network data not only describes the connections of network nodes but also describes the properties of the nodes. In this paper, we propose a community detection method collects relevant evidences from the information of node attributes and the information of network structure to assist the community detection task on node-attributed networks. We find communities in the framework of the semidefinite programming (SDP) method. In practical applications, the distribution of some node attributes may be uncorrelated with the network structure or the network itself may contain no communities as in a random graph. A sparse attribute self-adjustment mechanism is introduced to determine the relative importance of each source of information. As a by product, our method is also effective for community detection of multilayer networks that allow for multiple kinds of relations over the same set of nodes. Experimental results demonstrate the effectiveness of the proposed method.