Relatively Important Nodes Mining Method Based on Neighbor Layer Diffuse

At present, there have been some achievements in mining methods for relatively important nodes in complex networks, but the efficiency and accuracy of the methods still need to be improved. Based on the assumption that if a node has more neighbor nodes with certain characteristics, the more likely this node has such characteristics. This paper proposes a relatively important node mining algorithm based on neighbor layer diffuse (NLD), and verifies the accuracy and applicability of the method through experimental comparison and analysis.