Network Community Detection Algorithm for Graph Networks

This chapter presents how communities in networks can be detected by integrating barycentric serialization with bottom-up segmentation. Because nodes are efficiently ordered according to their neighbors by barycentric serialization, the segmentation algorithm provides modules in a computationally more efficient manner than the most frequently used Louvain community detection algorithms. The approach ensures efficient community detection by merging adjacent nodes or segments in a way that maximizes modularity, eliminating the need to test the entire dataset, and thus reducing iteration costs. Furthermore, the method is capable of accurately determining the number of communities in a network. The efficiency of the method is compared with other community detection algorithms based on benchmark problems.