Clustering dynamics of complex discrete-time networks and its application in community detection

The clustering phenomenon is common in real world networks. A discrete-time network model is proposed firstly in this paper, and then the phase clustering dynamics of the networks are studied carefully. The proposed model acts as a bridge between the dynamic phenomenon and the topology of a modular network. On one hand, phase clustering phenomenon will occur for a modular network by the proposed model; on the other hand, the communities can be identified from the clustering phenomenon. Beyond the phases' information, it is found that the frequencies of phases can be applied to community detection also with the proposed model. In specific, communities are identified from the information of phases and their frequencies of the nodes. Detailed algorithm for community detection is provided. Experiments show that the performance and efficiency of the dynamics based algorithm are competitive with recent modularity based algorithms in large scale networks.