Research On User Behavior In Social Networks Based On Evolutionary Games

Cooperative behavior has been widely existed in social networks,and information dissemination,as a kind of cooperative behavior,has been integrated into people’s daily social activities and influenced the development of network and human society.Therefore,the research of user cooperative behavior and information dissemination in social networks has become a hot issue in social network analysis.However,the influence of user behavior diversity and network complexity and dynamics on cooperative behavior has been neglected in the current research process,and most of the research on information dissemination has not considered the micro level.To this end,in response to the above-mentioned problems,this paper mainly accomplishes the following work.Firstly,the EEG(Edge Evolution Game)model is proposed for the characteristics of user behavior diversity.The model adopts a pairwise interaction mode,designs a income function based on the prisoner’s dilemma game psychology,and designs a strategy update rule based on the calculation results of users’ income;considering that the strategy update process affects the relationship between users,the disconnect-reconnect mechanism is proposed;the evolution of cooperative behaviors in different networks is analyzed through experiments,and it is verified that the disconnect-reconnect mechanism will promote the formation of cooperative clusters,thus promoting cooperation among users.Secondly,the ID-EEG(Information Dissemination based on Edge Evolution Game)model is proposed on the basis of the EEG.The model analyzes the information dissemination process at the micro level from the perspective of evolutionary game.The model divides users into infected and susceptible states,and uses cooperation and betrayal strategies to represent users’ behavioral intentions.Considering the different income obtained by different types of nodes after successfully disseminating information,information income is defined and a income function is designed;the probability of users receiving information is constructed according to the user incomes,and the dissemination behavior of users is simulated.Experiments were conducted on real network data,and it was found that information income would promote information dissemination,while betrayal temptation would inhibit information dissemination;it was verified that the ID-EEG model could accurately predict the information dissemination process and was stable.Finally,the ID-EEG is extended to construct MSID-EEG(Multi Source Information Dissemination based on Edge Evolutionary Games)model,which investigates the competition and cooperation relationship between information in social networks.By introducing influence coefficients,the relationship between messages is quantified in combination with information income;the dissemination process of competitive and cooperative messages is simulated,and the influence of influence coefficients on the dissemination process of cooperation and competition is analyzed.The experimental results show that: information in cooperative relationship has stronger facilitation ability as the influence coefficient increases;information in competitive relationship,the inhibitory effect increases as the influence coefficient decreases.