An Effective Model and Algorithm for Community Detection in Social Networks

In the research area of community detection in social network, there exist problems such as some algorithms with comparatively satisfactory detection result having high time complexity, current fast algorithms for large scale network resulting in low quality partition results, and lacking of model and mechanism to express and utilize actor and link attributes. To solve these problems, this paper proposes a model of edge stability coefficient and a model of complete information graph that can express the tightness of relation among actors, based on which an effective community detection algorithm is designed and implemented. The proposed model of complete information graph has high generality which makes it applicable to different community detection algorithms that need the input of fused information from actor and link attributes. Experiments show that the algorithm based on models of edge stability coefficient and complete information graph is effective to the problem of community detection in social networks with relatively less time cost. The algorithm is applicable to both weighted and unweighted networks with a comparatively fast speed and high quality partition result as well.