Detection of Influential Nodes Using Neighbor Closeness in Complex Networks

Identifying the influential nodes is important for understanding and controlling the dynamic processes, such as epidemic spreading, opinion dynamics, and cascading failures. Considering that the influence of nodes is strongly correlated with the degree, the first-order neighbors, and the second-order neighbors, we introduce a new measure, namely neighbor closeness (NC), which determines the influential nodes by calculating the neighbor closeness. To understand the accuracy of different methods and identify the influential nodes by low-complexity, we first study the identification of the influential nodes in the network based on different centrality metrics and take the mutual impact of the first-order and the second-order neighbors as the main factor to measure the influence of nodes. We use the Susceptible-Infected-Recovered (SIR) model and Kendall tau correlation coefficient to evaluate the performance and accuracy of the NC method, respectively. We find that there is a strong similarity between the NC method and the SIR model. Simulations on four real-world networks show that NC is an effective method to detect the influential nodes.