【摘要】 在线社交网络平台中存在的冲突信息加剧了群体观念对立，需要科学合理的管理机制来实现舆情治理。提出面向舆情治理的信息管理机制演化博弈模型，通过建立动态奖惩策略，自动化地实现冲突抑制，激活网络自我净化能力。通过引入演化博弈理论，构建基于累积前景理论的社交网络用户信息冲突演化博弈模型，使用数值模拟的方法研究9个约束条件下的18种理论应用情景的演化结果，分析了双方行为互动演化规律，得出约束条件五中博弈双方理性对待信息冲突的结论。通过建立博弈分析的系统动力学模型，对动力学演变进行分析，并在满足约束条件五的前提下对模型参数进行了详细的灵敏度分析，并为动态信息管理机制提供了实用化的策略建议。所提模型首次对利益驱动、从众心理等多种因素在舆情演化中的相互作用进行建模，演化分析结果与谣言优势传播规律相符合，证实了所提模型的合理性。所提模型为社交网络中恶意信息控制系统的开发和设计提供了重要借鉴，为舆情治理的研究实践提供了新的思路。更多还原

【Abstract】 Conflicting information in online social networks, often stemming from opposing views, can lead to rumors and the fragmentation of society. A reputation-based dynamic information management mechanism was proposed,adopting a dynamic reward and punishment strategy to decrease unnecessary information conflict so as to activate self-purification of network. By utilizing evolutionary game theory, a cumulative-prospect-based evolutionary game model of social network information conflict was constructed to validate the efficacy of the proposed mechanism. Numerical simulations were conducted to analyze the evolutionary trends of 18 different theoretical application scenarios under 9 constraints. The results demonstrate that, under constraint 5, both sides of the game tend to exhibit relatively rational behavior towards conflicts. Furthermore, dynamic evolution analysis and sensitivity analysis of the model parameters were performed while satisfying constraint 5, providing practical strategic suggestions for the dynamic information management mechanism. This model represents a novel attempt to capture the interaction of multiple factors, such as interest-driven behavior and conformity psychology, in the evolution of public opinion. Moreover, the model successfully simulates well-known patterns of rumor dissemination, confirming its validity. The proposed model offers valuable insights for the development and design of malicious information control systems in social networks and provides new ideas for the research and implementation of public opinion governance.更多还原