摘要
[目的/意义]探索构建网络舆情生态图谱应用于舆情演化过程的多维特征分析,为高效运作的舆情监管提供支撑。[方法/过程]从信息生态视角出发,以实体抽取及关系构造技术为基础,分别从舆情主题、参与主体、时序演化三个维度展开舆情生态图谱构建及多属性特征分析。[结果/结论]意见领域群体偏差易导致舆情衍生,主题特征挖掘及归因分析可实现衍生舆情的有效监测;影响力节点呈现出多元化与官方主导性,通过互动关系发现推动舆情发展的关键主体有助于合理调节舆情的聚合分化与演化方向;时序演化分析能有效提取舆情反转规律并辅助于舆情监测研判。
[Purpose/significance]This paper explored the construction of an ecological map of online public opinion,and applied it to the multidimensional characteristics analysis of the evolution process of public opinion,so as to provide support for efficient public opinion supervision.[Method/process]From the perspective of information ecology,based on the entity extraction and relationship construction techniques,this paper carried out the construction of public opinion ecological map and multi-attribute feature analysis from three dimensions:public opinion topics,participating subjects and temporal evolution,respectively.[Result/conclusion]The group deviation in opinion field can easily lead to the derivation of public opinion,and the topic feature mining and attribution analysis can realize the effective monitoring of the derivation of public opinion.The influence nodes show diversification and official dominance,and the key subjects promoting the development of public opinion can be found through the interactive relationship,which is conducive to the reasonable adjustment of the convergence,differentiation and evolution direction of public opinion.Time series evolution analysis can effectively extract the reversal rule of public opinion and assist in the monitoring and judgment of public opinion.
作者
刘申奥
李楠
Liu Shenao;Li Nan(Institute of Science and Technology Information,East China University of Science and Technology,Shanghai 200237)
出处
《情报探索》
2024年第12期44-52,共9页
Information Research
关键词
信息生态
舆情图谱
LDA主题模型
社会网络
情感演化
information ecology
public opinion map
LDA topic model
social network
emotional evolution
