Identify influential nodes in network of networks from the view of weighted information fusion

The network of networks (NONs) is a case of multiplex networks, when mining key nodes in the network, the information between the various sub-networks needs to be considered. In this paper, a weighted information fusion (WIF) method is proposed to identify the influential nodes of NONs. We first divide NONs into many individual networks and then perform weighted fusion. In the process, relevant information of nodes is measured to construct the basic probability assignment (BPA) for every single network. Besides, by considering the topological structure of the network, the method of effective distance is used to describe the weight of each BPA. Finally, to measure the influential nodes of NONs, the information of all single networks is fused to obtain structural information of NONs through WIF method. More than that, the influential nodes of four real-world NONs (including Neuronal and Social two types) are measured by the proposed method, and the results are compared with other five methods, which shows that WIF method is effective in identifying the influence of nodes of NONs.