Research On Recognition Method For Vital Nodes In Complex Networks

Complex network is an emergent interdisciplinary,which has occupied a high degree of scientific research in various fields in recent years.In nature,the majority of real networks can be abstracted into complex networks,which are usually composed of nodes,edges and weight.Under the background of network size increasing and network topology change more complex,how to accurately evaluate the importance of nodes is an important subject of complex network research.The main research of this paper includes ranking and excavating vital nodes:1.Aiming at the problem that the existing sorting algorithm has high time complexity and with a simple sorting mechanism,we proposes an algorithm based on the expansion factor of nodes which is dedicated to finding bridging nodes with small degree centrality but located among different communities.These nodes have a unique location advantage in the speed and scope of information propagation.The experiments show that the ranking result has high recognition accuracy,and it can find some important nodes that are ignored by other algorithms.2.Aiming at the problem that the existing excavating algorithm has high time complexity and overlapping in influence spread,we proposes an algorithm based on the local information of nodes which could distribute the initial seed nodes in reasonable positions to avoid rich-club phenomenon in the process of influence diffusion.The experiments show that the seed nodes have strong spreading capability.Furthermore,the algorithm is with linear time complexity,so the operation time is extremely short.This paper focuses on the research of vital node recognition methods,which proposes the algorithm based on the expansion factor and the algorithm based on local information of nodes.Our algorithms are faster than most existing algorithms and the sorting and mining accuracy can meet or exceed the level of existing algorithms.Our work in this paper complements the system of important node recognition algorithms research and improved algorithm performance.Our study has positive theoretical significance in network information mining and the research results can be applied to social networks,biological information,power networks and other practical fields,which has high application value.