A novel centrality measure for identifying influential nodes based on minimum weighted degree decomposition

It has theoretical interest and practical significance to find out influential nodes which make the information spread faster and more extensive in complex networks. A variety of centrality measures have been proposed to identify influential nodes, while numerous of them are one-sided and may lead to inaccurate for identification. To overcome this issue, based on the defined minimum weighted degree decomposition, we propose a novel centrality method for identifying influential nodes by combining the local and global information. First, considering the local topological attribute of node and spread characteristic of neighbor nodes, the local influentiality is defined as the node's influence in the local range. Then, a weighted neighborhood coreness centrality is presented as the node's global influence capability by taking into account the potential impact of edges on information dissemination among nodes and position characteristic of node. Finally, taking the combinatorial centrality of local and global range as the final influence of node is more comprehensive and universally applicable. We use Susceptible-Infected-Recovered (SIR) model, monotonicity, Kendall's tau correlation coefficient and imprecision function to estimate the performance of our method. Comparison experiments conducted on 14 real-world networks indicate the effectiveness of the proposed method.