Complex Network Community Detection Based on Swarm Aggregation

Finding communities in complex networks is not a trivial task. It not only can help to understand topological structure of large scale networks, but also is useful for data mining. In this paper, we propose a community detection technique based on the collective behavior of swarm aggregation, where all nodes are arranged on a circumference and each of them is assigned a angle at a random. The angles are gradually updated according to node's neighbors angle agreement. Finally, a stable state is reached and nodes belonging to the same community are aggregated together. By repeating this process, hierarchical community structure of input network can be obtained. The proposed technique is robust and efficient. Moreover, it is able to deal with both weighted and un-weighted networks.