Research On Hotspot Information Diffusion Mechanism In Social Networks

The online social media has gradually replaced traditional media as the main channel of information dissemination for its booming development,which has greatly promoted the speed and breadth of information dissemination and expanded the influence of informat ion.Therefore,understanding the spread of hotspot information on the online social networks is of great significance for advertising and public opinion control in the market.However,the influences affected by interactions of person-to-person,of person-to-information,and of information-to-information on information dissemination mechanism are not completely clear and need further study.The main researches are as follows:The theory of emotional infection in psychology shows that,in the process of information dissemination,people's emotions will affect their behaviors,thus affecting the information dissemination.In order to study the influences of different emotions on information dissemination,this thesis proposes an emotion-based informat io n dissemination model.Firstly,in order to analyze the role of different emotions in the process of information dissemination,the model introduced emotion communication power to describe the influence of different emotions on the probability of informat io n dissemination.In addition,since the connection relationship between users is the basis of information dissemination,this thesis introduced the power of information disseminat io n to study the effect of different connection relationships on the process of informat io n dissemination.Then,the dynamic equations have been established based on the heterogeneous average field equations,and the burst threshold and propagation scales have been theoretically derived.Finally,the propagation tendency of information with different emotions and connection relationships has been simulated,and the validity of the model has been verified through experiments.Most of the existing researches on multiple information dissemination considered the relationship between information based on a single topic,and ignored the impacts of user attention on information dissemination.To address this problem,firstly,this thesis quantified the impacts of user attention span on the user's ability to process informat ion,and classified information into different types according to the information contents relevance.Secondly,the nodes were divided into five states in combination with the SIR(Susceptible-infected-Recovered)epidemic propagation model,and the states transition matrix and its calculation formula have been derived by the discrete-time micro Markov chain theory.Finally,the non-linear dynamic model considering the interaction rules between nodes has been proposed by statistical methods.Equilibrium points of the model have been solved and its stability has been proved.This thesis analyzed the relationships between the user's final propagation probability,and the user's initial propagation wil ingness,and the number of received information paths,and the user's attention spans by numerical simulation.The competition or corporation relationships and the influe nces of attention spans on the information dissemination process have been analyzed by simulation experiments in the artificial network.The accuracy and validity of the model are verified by the fitting degree of real network data.