Research On Identification Of Influential Nodes Via PageRank And Spreading Probability

The progress of complex network analysis theory provides a new perspective and rich tools.A large number of real-world systems can be expressed as network structure,through abstraction as a unified model after quantitative research and solve practical problems,the content spans the social sciences Humanities,Applied Mathematics,physical science,financial fields and other multidisciplinary fields.As an important branch of complex network research,identification of vital nodes has profound theoretical significance and wide application prospect,which attracts more and more researchers ' attention.Important node recognition is a little of special nodes which play an vital role in the topology and function of the network,which has important influence on the antidestruction,evolution,control and synchronization of the network.According to the current mainstream infectious epidemic processes model and corresponding evaluation methods,in order to improve the accuracy and resolution of node importance identification,by analyzing the advantages and disadvantages of different central methods under different transmission rates,as well as the propagation overlap of top nodes after nodes ranking,the research contents of this paper are as follows: First,the basic concept of complex network is introduced,and the current node sorting method and node influence evaluation method are summarized and summarized systematically.Secondly,the advantages and disadvantages of various sorting algorithms under different propagation probability and the changes of PageRank algorithms under different tendencies are found through experiments.A PageRank improved algorithm DHP is proposed to modify the probability transfer matrix in combination with propagation probability.Through the experimental evaluation on the SIR propagation model,the algorithm can obtain good accuracy and resolution in many network data experiments,and has better adaptability to the change of propagation probability.Thirdly,a new identification of vital nodes method is proposed,which is based on community detection using DHP algorithm to select Seed nodes in the community,through the influence propagation probability,and the direct use of DHP algorithm to select seeds on the whole network can avoid the overlap of influence,to obtain a better scope of transmission.Fourth,the challenges and problems to be solved in the identification of vital nodes are summarized.