Identifying influential nodes in complex networks based on improved local gravity model

Abstract

Influential node identification has long been a focal point for researchers. Existing methods primarily focus on the individual topological characteristics of the nodes, making it difficult to accurately identify key nodes within a network. This paper introduces an improved local gravity model (ILGM) that incorporates node position, paths, quantity and injection to evaluate the influence of each node. The ILGM further explores the topological characteristics of neighbouring nodes, incorporating path and quantity data from adjacent nodes. This enhancement significantly improves the accuracy of the algorithm’s results. Empirical evaluations conducted on five real-world networks and one artificial network demonstrate that the proposed model effectively identifies influential nodes in complex networks.