Identifying influential nodes in complex networks based on resource allocation similarity

With the shift in the focus of network science research from macroscopic statistical regularities to microscopic scales, identifying influential nodes in networks has become a commonly discussed and challenging problem in network science. There has been substantial research on identifying the influential nodes, but most methods generally suffer from the incompatibility between time complexity and computational accuracy. This study aims to alleviate the contradiction between time complexity and computational accuracy. Therefore, the novel method based on resource allocation similarity(RAS) is proposed to improve degree centrality by creatively introducing a virtual node and combining it with similarity theory. We simulate the epidemic spreading experiment based on the Susceptible–Infected (SI) model, the static attacking experiment, and the node differentiation experiment on six classical networks. Four typical methods are used to compare with our methods. The experimental results show that, in most cases, the proposed method has a dominant advantage over other comparison methods.