Identification of important nodes in complex networks based on node and edge information

Aiming at the previous definition of important nodes, important nodes are redefined from the perspective of resource allocation and a new algorithm of identifying important nodes, DSE (Effective Distance-Structural Hole-Edges), is proposed. This algorithm utilizes the effective distance to represent the resource allocation, combines the number of ”holes” (i.e., the number of intermediary positions) and the importance of edges (where the K-shell algorithm is extended from identifying nodes to identifying edges). The performance of DSE and six commonly used centrality indicators in identifying important nodes is analyzed by using the SIR model simulation experiments in the nine real networks and the constructed networks. Finally, the feasibility of DSE algorithm is analyzed by using the Kendall’s tau correlation coefficient and the M(R) index. The experimental results show that the top-1% nodes and top-5% nodes identified by DSE have almost the highest ability to infect nodes in both the nine networks and the constructed networks. Additionally, the Kendall’s tau correlation coefficients are all positively correlated, and the M(R) values are all above 0.99. These show that the definition of important nodes in this paper is realistic and the DSE algorithm can effectively evaluate the importance of nodes.