Node Importance Recognition Algorithm Based on the Correlation Degree Transmission Contribution Matrix

Abstract

Based on the importance identification of nodes in a directed weighted network, a node importance recognition algorithm based on the correlation transmission contribution matrix is proposed. The proposed algorithm comprehensively considers the transmission ability of nodes, the correlation degree between nodes and adjacent nodes or indirectly related nodes, and the contribution degree between nodes to construct the correlation degree transmission contribution matrix. This paper combined the algorithm with the node efficiency of the local range of node importance and finally combined it with the global importance index intermediary number centrality to comprehensively evaluate the importance of nodes in the network. For the nodes with only incoming edges in the network, the product of the node mediator centrality and the correlation degree contribution matrix is calculated as the node importance. For nodes with only outgoing edges in the network, the product of node mediator centrality, degree value and node self-efficiency is calculated as node importance. Conducting experimental simulation analysis of the ARPA directed-weighted networks shows that the deliberate attack based on the algorithm results hits the network more obviously, which further proves the effectiveness and accuracy of the proposed algorithm.