K-clique community detection based on union-find

Abstract

As network community becoming increasingly complicated, the effective and fast community detection algorithm gets more important in network analysis. In this paper, a improved k-clique detection algorithm based on union-find structure is proposed, the time efficiency of community discovery in highly overlapped complex network is improved and it is possible to divide all communities within approximately linear time complexity. In this algorithm, we use union-find structure to store divided communities and reduce the number of unnecessary intersection test. The experiments result on the real data sets show that the algorithm is reasonable and effective, and its time efficiency is better than other overlapping community algorithms.