摘要
[目的/意义]基于突发公共卫生事件期间由公民隐私泄露导致的舆情事件,构建公民隐私泄露舆情的情感演化图谱可以呈现突发公共卫生事件期间网民的情感演化特征,为舆情监管和舆情引导提供参考。[方法/过程]结合文本词语加权方法“词频—逆文档频率”(TF-IDF)的LDA主题挖掘、机器学习的情感分析和社会网络分析方法,基于舆情生命周期的不同阶段,构建突发公共卫生事件中公民隐私泄露的情感演化图谱分析模型。并以新型冠状病毒肺炎疫情期间“成都确诊女子隐私泄露”事件为研究样本话题,分析不同舆情阶段的主题挖掘和不同舆情阶段的情感演化图谱。[结果/结论]网络暴力、隐私泄露和疫情防疫是疫情期间隐私泄露舆情主要关注点,公众讨论具有交互式特征和不同舆情阶段内的多元化特征。
[Purpose/significance]Based on the public opinion event caused by citizens’privacy leaks during public health emergencies,this paper construct an emotional evolution map which can better present the emotional evolution characteristics of users and thus provide better guidance for public opinion guidance.[Method/process]This paper uses the LDA topic model,combined with TF-IDF topic generalization,machine learning sentiment analysis and social network analysis methods to build a sentiment evolution mapping model of public opinion on privacy leakage of public health emergencies based on different stages of public opinion life cycle.Using the“privacy leakage of a woman diagnosed in Chengdu”incident during COVID-19 as a sample topic,we conduct the topic mining and sentiment evolution mapping of different public opinion cycles.[Result/conclusion]The results of the study show that online violence,privacy leakage and epidemic prevention are the main concerns.Public discussions are interactive and diverse in different public opinion cycles.
出处
《情报理论与实践》
CSSCI
北大核心
2022年第3期19-27,共9页
Information Studies:Theory & Application
基金
国家社会科学基金重大项目“大数据驱动的社交网络舆情主题图谱构建及调控策略研究”的成果,项目编号:18ZDA310。