融合贝叶斯网络与结构建模的衍生舆情形成路径及实证研究OACHSSCD

[Purpose]This paper intends to take derivative public opinion as the research object,analyze the factors affecting its formation and the relationships between these factors,explore its formation path,and verify the effectiveness and feasibility of relevant analytical methods.[Method]This paper adopts a method integrating Bayesian networks and structural modeling techniques to analyze the formation path of derivative public opinion,and verifies the effectiveness and feasibility of this method through comparative experiments and empiri-cal analysis.[Result/Conclusion]The research results show that:The formation path of derivative public opinion has been successfully identified,that the secondary effects have been quantified,and the causal hierarchical relationships of key variables have been systematical-ly revealed;The accuracy of this formation path in identifying the types of derivative public opinion formation reaches 93.33%,which verifies the explanatory power and robustness of the model in the multidimensional nonlinear public opinion interaction mechanism;Through Bayesian networks,it is identified that the highest number of reposts in a single day,the total number of reposts,the degree of public harm of the topic,the sensitivity of the topic,and the total number of comments are the key variables in the formation process of de-rivative public opinion;Restricting reposting behavior can increase the probability of unformed derivative public opinion by 80%,which verifies the effectiveness of the intervention strategy.