Label Propagation Algorithm Based on Random Wandering for Community Detection

Community structure is an important property on complex networks, and finding communities in the field of social network analysis is of greater information value. The label propagation algorithm is widely used to detect communities on networks due to its advantage of near linear time complexity. However, the randomness of the LPA algorithm in the label selection process may lead to the detection of poor quality communities. Therefore, we propose a label propagation algorithm based on random wandering and updates the label propagation rules while taking the network information obtained from random wandering into account in our algorithm. The final experimental results show that the proposed algorithm detects communities satisfactorily on real networks.