Community Detection And Evolution In Complex Network Based On Similar Tightness

Abstract:The research of complex networks is an important issue in many domains and disciplines. Numerous studies have shown that different complex networks have many common structural features in the real world, such as small world property, no scaling and community structure. The field of community detection has attracted a lot of interest considering community structures as important features of real-world networks. Community structure of the network has important practical significance to understand the network and analyze network characteristics. Community detection received an increasing interest as a way to uncover the structure of networks by grouping nodes into communities more densely connected internally than externally.The existing community detection algorithm is based on the density of nodes and edges without considering similarity among node objects, but in the real network the probability that similar nodes are in the same community is higher. In this thesis, it is proposed that a method of community detection is based on the local optimization of a similarity tightness function. According to nodes can be shared between different communities or not, the algorithm can find both nonoverlapping and overlapping communities. As applicable to the dynamic network, community evolution algorithm based on event has been proposed. It takes different operations to update the local community structure according to the events’ type. The similarity tightness function is used to update the local community structure.Algorithms are tested on real-world datasets and Synthetic benchmark networks. It is shown that community detection algorithm based on local similarity tightness is efficient and well-behaved. The community evolution algorithm based on event has low time consumption.