互联网舆情演化中群体行为协同演进模型研究

The evolution of internet public opinion is the result of interactive behavior of the broad group of net citizens, the process of which interweaves topic spreading and the evolution of net citizens' views on the public opinion, and both are interaction, which shows the more radical characteristics of dynamics.In order to describe the synergistic relationship between these two collective behaviors, the co-evolution model of collective behaviors in the evolution of the internet public opinion is proposed which includes the topic propagation sub-model based on the immune degradation strategy, the opinion evolution sub-model based on the dynamic networks as well as the co-evolution mechanism of both. The model considers the influence of both the net individuals and media on the behavior. Based on this model, the dynamic characteristics of the topic spreading and the opinion evolution behavior under the collaborative environment, and the principle of the interaction between the two are analyzed. At last, based on the above conclusion and combined the experiment simulation, the guidance and control strategy of irrationally collective behavior of the government is studied. The significance of research is that the co-evolution model of collective behavior breaks through the ideas and framework of the present work which studies topic spreading or opinion evolution behavior only. Furthermore, this model describes more realistically the mentality of net citizens and the evolution process of collective behavior. And it is also important for improving the capacity of the government to early-warn and handle the internet public opinion timely.

互联网舆情演化是广大网民群体行为交互作用的结果，演化过程交织着舆情话题传播和网民对舆情的意见演进行为，且两者相互作用，表现出更为激进的动力学特征。为了刻画出舆情演化中群体行为间的协同作用关系，在综合考虑网民个体和媒体对行为影响特性的基础上，研究舆情演化中群体行为的协同演进模型，包括：基于免疫退化策略的话题传播行为子模型、基于动态网络的意见演进行为子模型，以及话题传播和意见演进行为的协同演进机制；基于此分析协同环境下话题传播和意见演进行为的动力学特性，以及两者的相互作用原理；最后，基于群体行为动力学仿真分析的结论，结合实验仿真，研究政府对舆情演化中非理性群体行为的引导和控制策略。其意义在于，舆情演化中群体行为协同演进模型突破了目前工作只是单独针对话题传播或者意见演进行为进行研究的思路和框架，更为科学地刻画出网民心理和行为演进过程，对于提高政府对互联网舆情的预警和及时处理能力具有重要意义。

互联网舆情的演化是个复杂过程，对舆情的演化趋势进行有效监控和预警，是网络这一新兴媒体所面临的现实课题与严峻挑战。舆情的演化过程交织着广大网民对话题的传播行为和对舆情的意见演进行为，且两者相互作用，表现出激进的动力学特征。本项目着重研究了舆情演化中话题传播和意见演进行为形成及其协同演进的内在机理和发展规律，为政府监管部门加强对舆情的管制提供科学依据。首先，运用实证和实验仿真方法分析了网民、政府、媒体和网络结构等因素对互联网舆情演化中群体行为的作用及其影响特性；其次，针对舆情演化中的话题传播行为，基于复杂网络理论、传染病动力学模型和演化博弈等理论，分别考虑网络结构特性、话题衍生性、网民个体属性与媒体等因素对话题传播过程的影响，构建了互联网舆情话题传播模型，并通过实验仿真方法分析了舆情话题传播的动力学特征；第三，针对舆情演化中的意见演进行为，分别基于静态与动态网络结构，构建网民次近邻个体、媒体和环境噪声等因素影响下的舆情意见演进模型，并通过实验仿真方法分析了舆情意见演进的动力学特征；第四，为了描述舆情演化中话题传播和意见演进行为的协同作用过程，研究了媒体、政府影响下的话题传播和意见演进的动态交互机制以及两者协同演进的模型，并仿真分析了话题传播和意见演进的协同作用特性及其动力学特征；最后，基于模型的仿真结果探讨了政府对舆情的引导策略，并从网络科学角度，提出了社交网络中具有重要传播价值的关键节点、桥节点、弱连接路径的识别方法，为互联网舆情演化中群体行为的引导和有效干预提供了理论依据。项目研究科学地刻画出互联网舆情演化中网民心理和行为的演进过程，对于提高政府对舆情的预警和及时处理能力具有重要意义。