Evidential identification of influential nodes in network of networks

Abstract

Network theory has been widely used to describe complex systems in the real world. However, following the research of networks science, a network with one relationship in it can not be used in a multi-relationship complex system. Therefore, new network models were proposed to describe different relationships between real systems, such as Multilayers Networks, Multiplex Networks, Interconnected Networks and Network of Networks (NON). One type of NON has the same nodes but different relationship between nodes in different layers. This work focuses on this type of NON.In this paper, a new method is proposed to identify the influential nodes in the NON mentioned above based on evidence theory. Depending on the new method, influence of each node in different layer is fused by combination rules of evidence theory. Thus, results of fusion are used to quantify the influence of each node in NON to identify the influential nodes. Here we use the China transport networks, which can be regarded as a NON, to test the new method. The results of experiment show that the new method has smaller information loss during calculation and it is a reasonable method to identify influential nodes in a NON.