摘要
[目的/意义]基于舆情分析和仿真分析相关技术,耦合历史数据影响,创新性地探索多元主体协同治理模式干预舆情走势的最佳时间点,为政府治理相关舆情提供依据。[方法/过程]以延迟退休政策相关话题文本为例,借助BERTopic模型挖掘政策利益相关者关注热点,并基于ELMo-CNN模型绘制舆情情感演化图,将政策舆情演化阶段划分为扩散、高潮、维持和消退4个阶段。基于SD+ABM耦合模型,在考虑舆情演化各阶段历史数据依赖性前提下,仿真分析在4个阶段开始节点介入“政策—政府+平台(媒体)+公众”多主体协同治理模式下的舆情演化规律。[结果/结论]将仿真结果与原始舆情演化进行对比,探究出在扩散与高潮阶段进行多元主体协同治理干预效果最佳。
[Purpose/significance]Based on the relevant technologies of public opinion analysis and simulation analysis,coupled with the influence of historical data,this paper innovatively explores the optimal time points for the multi-subject collaborative governance model to intervene in the trend of public opinion,providing a basis for the government to govern related public opinions.[Method/process]Taking the topic texts related to the delayed retirement policy as an example,this paper uses the BERTopic model to mine the hotspots of concern of policy stakeholders,and draws the evolution graph of public opinion sentiment based on the ELMo-CNN model.The evolution stages of policy public opinion are divided into four stages:diffusion,climax,maintenance and decline.Based on the SD+ABM coupling model and considering the historical data dependence of each stage of public opinion evolution,this paper simulates and analyzes the evolution law of public opinion under the multi-subject collaborative governance model of“policy-government+platform(media)+public”intervened at the beginning nodes of the four stages.[Result/conclusion]This paper compares the simulation results with the original evolution of public opinion and explores that the intervention effect of multi-subject collaborative governance in the diffusion and climax stages is the best.
作者
方馨
刘高峰
龚妍婷
徐绪堪
Fang Xin;Liu Gaofeng;Gong Yanting;Xu Xukan(School of Business,Hohai University,Nanjing Jiangsu 210098;Institute of Management Science,Hohai University,Nanjing Jiangsu 210098;Institute of Statistics and Data Science,Hohai University,Changzhou Jiangsu 213022;Changzhou Key Laboratory of Industrial Big Data Mining and Knowledge Management,Changzhou Jiangsu 213022)
出处
《情报探索》
2026年第1期51-59,共9页
Information Research
基金
国家社会科学基金重大项目“总体国家安全观下重大突发事件的智能决策情报体系研究”(项目编号:20&ZD125)成果之一。
关键词
政策舆情
舆情演化仿真
SD+ABM耦合模型
多元主体协同治理
policy public opinion
simulation of public opinion evolution
SD+ABM coupling model
multi-subject collaborative governance
