学报首页 >> 期刊浏览 >> 正文

 标    题: 融合主题与细粒度情感特征的气候变化微博舆情分析研究(2023 第2卷 第4期 9)
作       者:

王君泽 詹若贤 李怡 杜洪涛

文章栏目:

数字经济与管理

摘       要:

我国在应对气候变化方面承担着重要的责任。研究公众对气候变化的关注程度和态度对于了解民意倾向、制定相应政策以及提升公民参与气候变化行动的意愿至关重要。本文在收集了2018年7月16日到2023年6月12日期间气候变化微博舆情数据,通过LDA主题模型和基于ERNIE预训练语言模型实现的情感细粒度分析,讨论了气候变化舆情中的公众关注热点、主题用户群体特征和情感倾向变化。研究发现,微博上关于中国在应对气候变化方面的重要作用以及全球气候变化的影响和相关科学研究讨论数量占比最大。人们还关注新技术的应用、气候变化与人体健康的关系以及明星公益行动等话题。博文的中立情绪占比相对较多,机构用户基本主导了话题,可能存在公众缺乏实质性参与的问题。气候变化认知主题下的博文负面情绪占比相对较高,该主题中的政府参与也有待提高。政务部门应当加强科学传播,提高公众对气候变化问题的认知和意识,增强公众的参与感。最后,本文仅考虑了博文内容和用户群体的特征,后续可以引入地域因素,并考虑结合文本和图片对多模态数据进行分析。

关  键  词:

气候变化;网络舆情;社交媒体;文本挖掘

Abstract:

China bears an important responsibility in addressing climate change. Studying public concern and attitude towards climate change is crucial for understanding public opinion trends, formulating corresponding policies, and enhancing citizens' willingness to participate in climate change actions. Collecting the microblogs of climate change from July 16, 2018 to June 12, 2023, this paper discusses the hot spots of public concern, the characteristics of user groups in different topics and the emotional tendency through LDA topic model and emotional fine-grained analysis based on ERNIE pre-training language model. The research results indicate that the number of discussions on China's important role in coping with climate change and the impact of global climate change and related scientific research accounts for the largest proportion. In addition, people are also paying attention to topics such as the application of new technologies, the relationship between climate change and human health, and celebrity public welfare actions. There is a high proportion of neutral sentiment in blog posts, and institutional users basically dominate the topic, which may lead to a lack of substantive public participation. The proportion of microblogs' negative emotions under the topic of climate change awareness is relatively high, and the government participation in this topic also needs to be improved. Government departments should strengthen the dissemination of science, improve the public's awareness of climate change issues, and enhance the public's sense of participation. Finally, this paper only takes into account the characteristics of blog content and user groups. Subsequently, regional factors can be introduced, and multimodal data can be analyzed by combining text and images.

Keywords:

Climate Change; Public Opinion; Social Media; Text Mining

作者简介:

王君泽,18luck新利电竞 公共管理学院,副教授,Email:wangjunze @mail.hust.edu.cn;詹若贤,18luck新利电竞 公共管理学院硕士研究生,Email:bxfz @vip.qq.com;李怡,18luck新利电竞 公共管理学院硕士研究生,Email:2689978102 @qq.com;杜洪涛,国家工业信息安全发展研究中心系统所,副研究员,Email:du14@tsinghua. org.cn。

链       接: 阅读原文




分享新闻:
0

 

 友情链接