Research On Consensus Formation In Internet Information Spreading Structure

With the fast development of Network and Communication technologies, Internet has become an important location where people gain information, express views and interact with each other. Different from traditional way of information spreading, Internet information possesses the features of time-space-free, free, bidirectional and productive. Traditional mathematical model of opinion dynamics obviously can not depict the property of information spreading and interpersonal interaction, as well as explain the phenomenon of Internet information spreading and opinion evolution. Hence, it began receiving an enormous amount of attention in the field of information, complex system, and interdisciplines, to investigate Internet information spreading structure, and its effects on opinion evolution and consensus formation.Based on the data and phenomenon of user discussion in Internet forum, the dissertation is devoted to statistically analyze the Internet information spreading structure, and it is found that this structure has modular structure. Then, the consensus formation in Internet information spreading structure is inspected. Especially, the opinion and inclination evolution with explicit community structure, the virtually social impact process in dynamic spreading structure with implicit coupled communities, and the interaction process of Internet users with bounded rationality is studied. The aim of dissertation is to understand the regularity of information spreading, opinion evolution and consensus formation on the Internet, and to discover the features of Internet information spreading. The work of the dissertation is supported by the National Natural Science Foundation of China (No.60972012), Beijing Natural Science Foundation (No. 4102047), Beijing Natural Science Foundation (No.4112045), Project supported by the Specialized Research Fund for the Doctoral Program of Higher Education of China (No. 20100009110002), the Academic Discipline and Postgraduate Education Project of Beijing Municipal Commission of Education, and the Fundamental Research Funds for the Central Universities (No.2009YJS007). Main contributions of the dissertation are as follows:1. Empirically analyze the statistical property of Internet information spreading structure. Construct the Internet information spreading structure based on implicit reply; reveal the character of Internet information spreading and interpersonal interaction. It is found that this structure is directional, asymmetry, disassortative, scale-free network, and has community structure. The empirical analysis is a foundation of the following research on consensus formation in Internet information spreading structure.2. Construct a model of network consensus formation with explicit community structure which is coupled with fixed nodes. Propose a bounded confidence model with opinion leader in coupled random networks, and a CODA model with global promotion in coupled regular networks. Make Monte Carlo simulation and find that, increasing interconnectivity can facilitate opinion transfer and sharing, promote the opinion convergence between different communities, and weaken the extremism.3. Propose virtual social impact model with implicit community structure due to Web encounter. Investigate the social impact and consensus formation in dynamic information spreading structure caused by encounter, and find that large encounter frequency can facilitate information propagation and opinion sharing, and promote the consistency in opinions between different communities. What's more, results show that the interface node is not the only way through which different communities interact with each other. Instead, Web encounter also processes the coupling effect. Accordingly, in information promotion and opinion spreading, the spreading effects of interface members are limited, and it is necessary to develop the spreading effects of all members.4. Model the bounded rationality Internet users with selective attention. Based on anonymity and freedom of Internet, the interpersonal interaction model matching the computer mediated information propagation mode more closely is presented, and the information spreading mode in online mass incidents is discussed. Considering the interplay between interpersonal topology and opinion, research discover the condition for appearance of extremists, uncover the relationship between selective attention and group polarization, and suggest the non-mandatory strategy quickening or slowing down the speed of extremalization.