Vulnerability Assessment for Directed Weighted Complex Network by Quantifying Importance of Community Nodes

Vulnerability Assessment for Directed Weighted Complex Network by Quantifying Importance of Community Nodes

연구 분야: Infrastructure

학회: 2023 4th International Conference on Computer Engineering and Application (ICCEA)

Vulnerability assessment of directed weighted complex networks is the prerequisite for network structure optimization. In the present work, a method based on the importance of community structure nodes is proposed. First, complex network communities are detected from networks using an improved-permanence-based method. Second, importance of nodes are quantified by calculating both internal and external community connectivity. Then, the nodes are sorted in order of importance and destroyed one by one. Based on the destroyed networks, indexes of network efficiency and connectivity are calculated, by which vulnerability of complex networks is assessed. The experimental results show that: 1) the proposed method can evaluate the vulnerability of directed weighted complex networks with multiple scales. 2) the proposed method shows good efficiency in network destruction. 3) in terms of the ability to completely destroy the network, the proportion of nodes to be destroyed is only 0.8261, 0.3200 and 0.8200 for test set, representing better performance of our method in accurately finding vulnerable nodes or regions.