Overlapping Community Discovery based on Seed Expansion and Shortest Path of Graph

Abstract

This article proposes an overlapping community discovery algorithm based on the combination of seed extension and the shortest path of the graph (Overlapping Communities Detecting Based on Seed Extension and Shortest Path of Graph, the following referred to as SESPG algorithm). First, the seed node is selected according to the influence of the node. Secondly, the nodes of the community are preliminarily divided by calculating the similarity between the nodes and the seed community. Then calculate the maximum value of the shortest path between the seeds in the community and the nodes in the community. Use this as the radius. For nodes within the radius but not in the community, calculate the similarity with the community, and join the community if it exceeds the threshold. Finally, the communities with high similarity are merged to complete the discovery of overlapping communities. After comparing with other algorithms, the SESPG algorithm has a good performance when the complexity is high, and the number of nodes is large.