Node importance evaluation method of complex network based on the fusion gravity model

The study of complex networks is increasingly attracting widespread attention, and the importance analysis of key nodes with significant influence has always been a core problem in network science. Currently, many centrality measures and gravity model methods usually only focus on a single attribute of a node to evaluate its importance. However, they often ignore the influence of node multi-attribute characteristics on the evaluation results. In order to analyze influential nodes in complex networks more effectively, the local and global information of nodes must be fully considered. To solve this problem, based on the multi attribute characteristics of nodes and the gravity model, we propose a new fusion gravity model that fuses the multi attribute characteristics of nodes to evaluate the influential nodes in the network more comprehensively. The model takes into account the local information of the network reflected by the node degree value, the global information of the network reflected by the weight of the maximum eigenvector of the node, and the location information reflected by the node K-shell value. Finally, we used the SI (Susceptible-Infection) propagation model to conduct simulation experiments based on six real network datasets, compared with the traditional centrality method and similar methods, and verified the rationality and excellence of the proposed method.