Abstract:

Particle swarm optimization (PSO) is addressed into community detection problem, and an algorithm based on voting strategy is proposed. In contrast with other label propagation strategies, the main contribution is to take the impact of node into consideration, in which not only the number of nodes with the same label in its neighbors, but also the degree of that node are considered. Special initialization and update approaches based on it are designed in order to make full use of it. Experiments on synthetic and real life networks show the effectivity of proposed strategy. Experiments on reallife networks also demonstrate it is an efficacious way to solve community detection problem.