Research on Public Opinion Monitoring and Evaluation Indicator System and Risk Stratification Based on Three-Dimensional Modeling

In today’s digital communication environment, the outbreak of public crisis events is often accompanied by the rapid spread of public opinion information and immediate feedback of public sentiment, forming a complex public opinion field. Traditional public opinion monitoring methods lack dynamic modeling capabilities, making it difficult to cope with the multiple challenges of the variability in the speed of public opinion dissemination, the heterogeneity of stages, and the complexity of emotional tendencies. To cope with this problem, a three-dimensional indicator system based on the dissemination time, scale and emotional tendency of netizens is proposed, and the weight parameters are determined by the hierarchical analysis method (AHP) and the Delphi method, which is combined with the three-dimensional visualization technology to dynamically track the trajectory of public opinion evolution and realize the four-level risk grading from slight to dangerous. The experimental results show that the model can effectively identify public opinion events of different risk levels. By dynamically adjusting the weights of the indicators, the model can adapt to the complex changes of public opinion events. In the future, we will further optimize the model parameters and combine it with machine learning and natural language processing (NLP) technologies to enhance the monitoring capability and explore its application value in multiple fields.