An Overlapping Community Detection Algorithm for Label Propagation Based on Node Influence

摘要

LPA algorithm is an effective community detection algorithm. based on the original algorithm, a new algorithm based on node importance, similarity between nodes, influence of adjacent nodes and label propagation strategy is proposed in view of the randomness and instability of overlapping community detection in LPA algorithm. First, the importance of nodes, the similarity between nodes, the influence of adjacent nodes are calculated, and the label set of nodes is generated according to the calculation results. Through the label update strategy, the attribution coefficient of each node to the community is calculated, and the updated label set is iterated continuously. The results show that the algorithm has near linear time complexity and can effectively improve the accuracy and stability of large-scale overlapping community detection.