The Related Research Of Complex Network Model Based On Link Prediction

With the evolution and development of network science by the stages of regular networks and random networks,the structure and property of networks also become more and more complex,and traditional network models can not describe the topology of current networks accurately.With the gradual improvement of the complex network theory,many new methods and tools have been proposed for the study of real network topological characteristics.Nowadays,there is considerable interest in the link prediction of complex networks,because it is able to predict the unknown links according to the information in the networks.In this paper,a real network model is proposed firstly,and the model is proved to have a stationary power law degree distribution.It is also verified that the clustering coefficient of the model is independent of the network size.Then for the currently popular WeChat circle of friends,we give some qualitative explanations to some phenomena by using this complex network model.Furthermore,we construct a model of link prediction based on probability method,and give a new model for the network by adding a predictor.We find that the new model has the same properties.In addition,we show that the model has a stationary degree distribution and high clustering coefficient when the network size tends to infinity by computer simulations and mappings.Thereby,it is theoretically and empirically proved that the model has practical significance.