Research On Key Node Identification In Directed Weighted Networks

In recent years,due to the rapid development of network technology,complex network has gradually become a hot spot in many fields.Most of the complex systems in real life can be abstracted as complex networks.In the research of complex networks,it is of great theoretical and practical value to identify the key nodes.At present,many methods for key node identification in complex networks have been proposed by scholars at home and abroad,but most of them are suitable for the study of undirected unweighted or undirected weighted networks,however,in most real-life complex systems,even the flow of information in the direction has to consider the weight,so it is necessary to propose the evaluation index for the key nodes of the directed weighted network.However,the existing evaluation indexes often only consider the characteristics of the nodes themselves,ignoring the influence of the neighbor nodes and other nodes in the network,the importance of nodes is only considered locally,while the whole structure of the network is ignored.In view of the above problems,this article has done the following several aspects of research,specifically as follows:First of all,through the research background and research significance of the paper carding research issues,with the help of complex networks in the knowledge of graph theory to describe different types of networks.A key node identification algorithm based on integrated strength and node efficiency for directed weighted networks is proposed,which fully considers the influence of the edge weight between nodes on the node importance,and the flow of local and global information.Secondly,in order to verify the validity and accuracy of the proposed algorithm,the complex network key node identification algorithm is applied to the directed weighted ARPA network and the directed weighted network with symmetric structure,taking the number of subgraphs and the maximum size of subgraphs as the evaluation criteria of node recognition accuracy,four different key node recognition methods are compared and analyzed.The experimental results show that the proposed method can distinguish the differences among the nodes and effectively identify the key nodes.Finally,in order to further validate and apply the proposed algorithm,this paper takes China’s high-speed railway network as an example to carry out experimental simulation analysis,and obtains the relevant network parameters of China’s high-speed railway network by using UCINET network simulation software,further use of MATLAB software to calculate the comprehensive strength of each node in the network and node efficiency,and then get the importance of different nodes in the network.At the same time,the practical significance of the key nodes in China’s high-speed railway network is analyzed based on the geographical location map and the functions of each node in the network.