A Novel Method of Influence Ranking via Node Degree and H-index for Community Detection

摘要

Identifying influential nodes is critical to have a better understanding of the network function and the process of information diffusion. Traditional methods of evaluating influential nodes such as degree centrality ignore the location of a node and its neighbors' influence in networks, while this plays an important role in revealing the node's local influence in spreading information. In this paper, we propose a novel method, named DH-index (node Degree and H-index), to measure a node' importance by considering its and neighbors' influence simultaneously. Meanwhile, we put forward a node DH-index based label propagation algorithm (DH_LPA) for community detection. We demonstrate its validity and feasibility on a set of real-world and synthetic networks for our new proposed community detection method.