Information Propagation Research Based On Coupling Of Neighbor Nodes

Today is an era of information explosion,countless information is generated and propagated every day,especially with the development of Internet technology and the application of self-media technology,the number of individuals involved in information propagation increases,and the scope of information propagation increases,at the same time,the way of information propagation is changing,previously,the public could only receive information from fixed media,such as TV,newspaper,etc.Nowadays,the public can obtain the information they are interested in through the Internet,so the factors affecting the propagation of information have become more diverse and more complex.Therefore,the study of information propagation,especially the study of special factors,has far-reaching significance.With more and more research results of complex network theory have been found,scholars in various disciplines have also applied complex networks to the corresponding fields.Many problems have new solutions,for example,the establishment of infectious disease transmission dynamics equations,the financial equity allocation problem,and the study of information and rumor transmission laws are all based on complex network theory.Based on the existing results,this paper studies the influence of several special factors on information propagation,and establishes a network model and a propagation model that better conform to the law of information propagation.The content of this article mainly includes the following two aspects.(1)Based on the ability of remember information,a network model with memory was constructed through multiple rounds of information dissemination training,and constructed an information dissemination model with cluster-selection ability based on unsupervised learning in machine learning.The memory of certain information will make people pay more attention to and more actively participation in the similar information dissemination when they encounter it again,which makes the information spread faster and the dissemination scope is wider.We conducted a comparison experiment through BA scale-free network model and network model with memory,and found that the dissemination effect of the network with historical memory information is better and more accord with reality.(2)The influence of surrounding neighbors is the most direct and powerful factor affecting the spread of information.We abstracted the neighbors influence as attraction,and funded that there is a close relationship between information transmission and neighbor nodes by studying the influence of neighbors in different states on the attracted individuals.In the real information transmission,not only the first layer neighbors have a direct influence on the information transmission,sometimes,the two-layer neighbors also play an important role in the information transmission.Based on the above analysis,we combined an information propagation model with the neighbor nodes attractiveness based on the critical value model and the Poisson model as an influence factor.We have not only studied the influence of the first-layer neighbor nodes gravity on information dissemination,but also extended to the two-level neighbor nodes gravity.By comparing the results of the simulation experiments with the disseminating results of two real cases of “the picture infringement of Visual China” and “Sri Lanka terrorist attack”,we found that the simulation results of the information propagation model containing the first-layer neighbor nodes gravity are consistent with the cases(“ the picture infringement of Visual China”),which have less people involved in information spread,and the simulation results of the information dissemination model containing the two-level neighbor nodes gravity are closer to the cases(“Sri Lanka Horror" Assault ”),which have a large number of participants in information dissemination.