An improved SGN algorithm research for detecting community structure in complex network

Abstract

In order to make more accurate partition community structure of complex networks, this paper puts forward a new community partition algorithm. The basic idea of the algorithm depends on node similarity, and it deletes the link whose similarity is the smallest every time, then takes modularity Q as the judging standard. Computing the corresponding modularity when network occurs into pieces, and the module structure is the ultimate community structure when Q reaches its peak. This algorithm not only improves the accuracy of the original algorithms, but also makes sure that the community structure has a better quantification. When the new algorithm is applied to the complex networks, we finally find that the algorithm is effective and feasible. © 2014 Springer International Publishing.