Study On The Evolution Of Collective Behavior In Public Opinion Propagation Based On Social Networks

Public opinion spread in the public through the social networks, when people establish contact through a variety of networks, they will affected by the behavior and decisions of other people, or the entire society circles of public opinion. The process of effect and effected is the process of cluster behavior evolved in the social networks. With various types of network group incidents has appeared in recent years, the behavior of the people affects the emergence and development of these events. The process of cluster behavior which evolved from individual behavior is generated by continuous contact, the impact of the transfer of information is the key to the process of dissemination of public opinion, and therefore, it’s easy to see that dissemination of public opinion plays an important role in the evolution of cluster behavior. Then, study on the evolution of cluster behavior must proceed from the spread of public opinion. In this paper, the general rules of the evolution of behavior cluster is revealed, which hope to provide some relevant theoretical support to understand the social state departments of public opinion accurately, grasp and manage public opinion and establishment some control laws and regulations.From complex social network theory, it’s obvious that research network behavior from the structure of the network is an effective way to analysis the rules of behavior. Microcosmic social network refers to the boundary and the order of the group, a macro perspective of social network is without boundary and order of the network, but the basis of the division is the social network of two hierarchies. In this paper, the research on the evolution of public opinion is divided into two aspects in social networks. On the one hand, take the network as a group of relatively unorganized individuals, and then examine the aggregate effect of the entire group. On the other hand, use the specific structure of the network to analysis that how individuals affected by other neighboring network nodes. Start from this two aspect, the basic process of spread public opinion in social networks is analyzed, and information cascading effect and structural effects of public opinion propagation network is also founded in this process. Modeling and analysis the evolution of cluster behavior which based on this two effect, the specific research and theoretical contributions are as follows:1) Studied the evolution of cluster behavior in the dissemination of public opinion from the group. Firstly, the cluster process of public opinion propagation was description and analysis. Secondly, combing the cluster behavior evolution model elements, using Bayesian rule were constructed of public opinion communication in cluster evolution behavior of individual behavior analysis model and set of group behavior influencing model. Thirdly, solve the model obtained cluster evolution behavior of critical conditions, analysis of the evolution of the collective behavior of the spread of public opinion information cascade effect. Thus, this paper puts forward the some suggestions for control of collective behavior, which share a good signal and released public signals help to steady the correct number of clusters behavior, eliminate the error cluster behavior.2) Studied the evolution of cluster behavior in the dissemination of public opinion from the social network. Firstly, from the social network perspective, the evolution of public opinion propagation is described. Secondly, from the two aspects of evolution of individual behavior and group behavior of the game, using the method of cluster coordination game behavior evolution model analysis. Thirdly, construct the cluster behavior in the dissemination of public opinion by solving the evolution model. Finally results of simulation model, the different factors and the structure of network effects in the quantitative analysis, simulation results show that the larger positive conversion rate and external effect rate, is conducive to the cluster communication in social networks, and put forward the control strategy of cluster behavior, namely through adjust the threshold value and return value can realize the control of the cluster behavior.