Research On Game Mechanism And Trust Prediction In Complex Networks

The emergence of complex network provides a new perspective and a new approach to study complexity,which describes the real-world networks with an abstract way.In the meantime,the introduction of evolutionary game theory offers a strong theoretical tool for us to solve some social problems,and the combination of complex network and evolutionary game theory deeply broadens the understanding of the evolution of cooperation.According to the theory of complex network,the system modeling and simulation is utilized to study the evolution of cooperation behavior on the prisoner's dilemma,snowdrift game,and public goods game model.In addition,the thinking of complex network is used to model trust prediction for online social networks.The main results and innovative points can be described as follows:(1)To propose a novel self-interaction mechanism,which is applied into the structured network to investigate the influence of evolution behavior of self-interaction in the prisoner's dilemma and snowdrift game.Based on the prisoner's dilemma and snowdrift game,a variety of forms of self-interaction strength are considered to investigate the effect of group cooperative behavior.Furthermore,the impact of individual difference on cooperative behavior,on the basis of self-interaction game mechanism is also studied here.The numerical simulation results clearly show that the introduction of the self-interaction mechanism can significantly improve the level of cooperation of structured groups.The research results can help to allocate all kinds of resources effectively,avoid the situation of inefficiency,and is helpful to solve the problem of credit construction.(2)To explore the influence of individual difference and heterogeneity of investment on the public goods game,in which two types of individuals,A and B type.It is supposed that the individual difference can be achieved by controlling the proportion of A-type individuals in the group,and ? parameter is used to adjust the degree of investment heterogeneity in PGG group.The simulation results show that the A-type individuals and the investment heterogeneity has a positive effect on the group cooperation,greatly enhance the level of cooperation.The current results offer a good theoretical framework and solution for people to solve the problem of providing public services and reduce greenhouse gas emissions.(3)To construct a new mathematical model for trust prediction in social networks based on the modeling thinking of complex network.Various social factors are incooperated into the model,and the unsupervised learning method is used to predict the trust relationship between users.Meanwhile,public data sets Epinions are used to test the present algorithm and initial results show that the introduction of homophily and time effect in the model can improve the prediction accuracy and enhance the scope of information spreading,which make the information transmission more effective in the online social networks.