Importance evaluation method of complex network nodes based on information entropy and iteration factor

Abstract

In the study of complex networks, scholars have long focused on the identification of influencing nodes. Based on topological information, several quantitative methods for determining the importance of nodes are proposed. K-shell is an efficient way to find potentially affected nodes. However, K-shell overemphasizes the influence of the location of the central node and ignores the influence of the force of the nodes located at the periphery of the network. Furthermore, the topology of real networks is complex, which makes the computation of the K-shell problem for large scale-free networks extremely difficult. In order to avoid ignoring the contribution of any node in the network to the propagation, this paper proposes an improved method based on iteration factor and information entropy to estimate the propagation ability of each layer of nodes. This method not only achieves the accuracy of node ordering, but also effectively avoids the phenomenon of rich clubs. To evaluate the performance of this method, the SIR model is used to simulate the propagation efficiency of each node, and the algorithm is compared with other algorithms. Experimental results show that this method has better performance than other methods and is suitable for large-scale networks.

Identifying influential spreaders in complex networks based on density entropy and community structure

“…Zhong et al [21] proposed an improved information entropy (IIE) method that considers both the number of neighbors surrounding the target node and its rate of spreading. Wang et al [22] proposed an enhanced method utilizing the iteration factor and information entropy to estimate the propagation capability of each layer of nodes. It not only accurately orders the nodes but also effectively avoids the phenomenon of rich clubs.…”