Research On Individual Decision-Making And Opinion Interacting Mechanism In Internet

Internet applications such as Social Networking Services have become an important source of online public opinion in the Internet age. With the rising popularity of the online applications, interactions among individuals have been easily carried out. The formation of online opinion depends on individual decision-making, information spreading and interactions among individuals, which form a complete generating system of online opinion. Different from traditional way of public opinion, online public opinion possesses new features such as strong timeliness, weak geographical restrictions and bidirectional. For this reason, traditional model and algorithm of social opinion can not describe the property of online opinion interaction accurately and holisticly. Since online public opinion has many new features, to research how such features influence the process of online opinion formation is of great significance. The dissertation is devoted to analyze the properties of individual decision-making, and it is found that individuals’opinion always have a tendcy of extremist. We inspected the opinion diffusion model based on individual decision-making. The process of individual decision-making, the influence of opinion diffsuion and the characteristics of gruop opinion intercation are studied. The work of the dissertation helps to understand the opinion interacting behavior of Internet users, provides a demonstration for investigating user behavior and makes a contribution to further research on prediction strategy of online public opinion.The work of the dissertation is supported by the National Natural Science Foundation of China (No.61172072,61271308), Beijing Natural Science Foundation (No.4112045) and Specialized Research Fund for the Doctoral Program of Higher Education of China (No.20100009110002).Main contributions of the dissertation are as follows:1. Analyze the influence of community structure on information receipt-accept rule, model the behavior of individual decision-making. Extend the improved Receipt-Accept Sample model to the process of online interaction and simulate the group’s opinion evolution in different situations. Simulation results show that when the influence of network topology is removed, the news or messages from mainstream media decide the trend of opinion and the process of group evolution in community expands this trend. With the behavior of group interacting, individuals’ opinion maintain a trend of extremist.2. Construct an opinion dynamic model based on Innovation Diffusion theory. Examine the influence of agent’s observation range, innovator location and other factors on the process of opinion evolution. We carry out simulation on small-world network and real SNS network topology to test the stability of the model. Numerical results show that with the obvservation probability rising, clustered innovators have greater probability of surviving and carrying out diffusion successfully, even though randomly spread innovators have almost equally effective persuasion speed.3. Propose an opinion formation model which takes an individual’s opinion transition probability into consideration. We study the influence of Internet information, quantify the rule of group interaction and carry out simulation. The model introduces different parameter distributions which meet the current situation which simplifies the simulation of opinion evolution. Simulation results show that if a message’s view stays below a certain threshold and its credibility takes a high value, it can win the trust of agents with higher awareness, leading the public opinion to support the message’s view in the beginning and then decide the opinion formation.4. Empirically analyze the performance of proposed model based on a real hot topic. By quantitative information of Internet message, we set network topology, information features and initial opinion distribution of the model, carry out numerical simulation and discusse the prediction proposal and promotion strategy of online public opinion. The results show that evolution results of the model can depict the opinion trends of hot topic effectively.