A new scheme for identifying important nodes in complex networks based on generalized degree

Abstract

The identification of important nodes is a fundamental problem of complex networks. To address this issue, the k-shell decomposition algorithm and its improvements has been presented. The classical K-shell algorithm fails to provide the promising results in practical situations because of the resolution and performance problems. Therefore, this paper proposes a new scheme named generalized degree decomposition (GDD) algorithm for the identification of important nodes. In our scheme, first, each edge is assigned with hidden weight to describe its potential contribution to the network. Then, the generalized degree of each node is calculated based on the weights of the edges connected the node and its neighbors. At last, each node is attached with a corresponding index according to its stripping stage in the network pruning process. A large number of experiments are performed in eight real networks, and the findings demonstrate that the proposed scheme optimizes the performance and resolution.