ConvGraph: Community Detection of Homogeneous Relationships in Weighted Graphs

This paper proposes a new method, ConvGraph, to detect communities in highly cohesive and isolated weighted graphs, where the sum of weights is significantly higher inside than outside communities. The method starts by transforming original graph into line apply convolution, common technique computer vision field. Although this was originally conceived optimum edge images, it used here edges identified their rather topology. includes final refinement step applied with high vertex density that could not be detected first phase. proposed algorithm tested on series synthetic graphs real-world export graph, performing well both cases.