被高估的“领袖”:基于网络舆情案例数据分析的群体意见演化研究
The Overestimated “Leaders”: A Study on the Evolution of Group Opinions Based on the Data Analysis of Online Public Opinion Cases
摘要: 目的/意义:网络舆情事件的发酵过程中有大量意见领袖和普通网民的参与,但是两者的互动关系尚未得到深入的分析。本研究揭示意见领袖和普通网民在互联网舆情里的关系。方法/过程:使用微博社交媒体平台“胖猫”事件进行案例分析,采集5月2日~5月28日相关事件微博文本数据;通过BERTopic模型对文本内容进行主题建模;利用百度情感分析API对文本进行情感计算;将信息发布者划分为不同主体;最后,从主题、情感和不同主体三个维度分析不同主体在舆情事件里的关系。结果:研究结果表明,在舆情关注度层面,意见领袖多聚焦事件新动态和性别议题,而普通网民更倾向于对事件进行感悟与反思。从主题情绪分布来看,在事件感悟与反思议题上,普通网民的情绪较意见领袖更为消极;而在男女性别议题中,意见领袖的情绪较普通网民更为消极。情感分析显示,意见领袖的正向发声,未能有效提升普通网民的正向情感占比。结论:根据舆情主题关注度、情感以及主题情绪分布呈现出的差异性来看,意见领袖对普通网民的影响相对有限。
Abstract: Purpose/Meaning: During the fermentation process of online public opinion events, there is extensive participation from both opinion leaders and ordinary netizens. However, the interactive relationship between these two groups has not yet been thoroughly analyzed. This study aims to reveal the relationship between opinion leaders and ordinary netizens in internet public opinion. Methods/Processes: Using the “Fat Cat” event on the Weibo social media platform as a case study, relevant event-related Weibo text data from May 2 to May 28 were collected. The BERTopic model was employed for thematic modeling of the text content, and the Baidu Sentiment Analysis API was utilized for sentiment computation of the text. The information publishers were categorized into different entities. Finally, the relationships between different entities in the public opinion event were analyzed from three dimensions: themes, sentiments, and different entities. Results: The results indicate that in terms of public opinion attention, opinion leaders tend to focus more on new developments of the event and gender issues, while ordinary netizens are more inclined to reflect on and contemplate the event. In terms of thematic sentiment distribution, ordinary netizens exhibit more negative emotions than opinion leaders in the theme of event reflection and contemplation, while opinion leaders show more negative emotions than ordinary netizens in the male-female gender theme. Sentiment analysis shows that the positive voices of opinion leaders have not effectively increased the proportion of positive sentiments among ordinary netizens. Conclusion: Based on the differences in thematic attention, sentiment, and thematic sentiment distribution, it can be concluded that the influence of opinion leaders on ordinary netizens is relatively limited.
文章引用:赵端男, 钱颖. 被高估的“领袖”:基于网络舆情案例数据分析的群体意见演化研究[J]. 运筹与模糊学, 2025, 15(3): 95-106. https://doi.org/10.12677/orf.2025.153143

参考文献

[1] 陈永刚, 孙卉垚. 互联网舆情研究[J]. 情报杂志, 2011, 30(S1): 85-88.
[2] 李吉虎, 潘旭伟. 融合主题的社交媒体舆情SEIR传播模型构建与仿真研究[J]. 智能计算机与应用, 2024, 14(4): 34-44.
[3] 魏宇航, 田园. 高校突发事件网络舆情传播影响因素研究[J]. 情报科学, 2021, 39(12): 53-59.
[4] 胡凯茜, 李欣, 王龙腾. 基于BERTopic模型的网络暴力事件衍生舆情探测[J]. 情报杂志, 2024, 43(7): 146-153.
[5] 安璐, 吴林. 融合主题与情感特征的突发事件微博舆情演化分析[J]. 图书情报工作, 2017, 61(15): 120-129.
[6] Pi, Z.X. and Feng, H. (2022) The Evolution of Public Sentiment toward Government Management of Emergencies: Social Media Analytics. Frontiers in Ecology and Evolution, 10, Article ID: 1026175.
[7] Liu, Z., Wu, J., Wu, C.Y.H. and Xia, X. (2024) Shifting Sentiments: Analyzing Public Reaction to COVID-19 Containment Policies in Wuhan and Shanghai through Weibo Data. Humanities and Social Sciences Communications, 11, Article No. 1104. [Google Scholar] [CrossRef
[8] 代婧琦, 赵晔, 李宗敏. 情感与主题协同分析的突发事件舆情图谱构建[J]. 武汉理工大学学报(信息与管理工程版), 2023, 45(2): 224-233.
[9] 于凯, 王云乐. 互动仪式链视角下隐私泄露事件网络舆情分析[J]. 西南大学学报(自然科学版), 2023, 45(8): 48-60.
[10] 郑傲. 网络互动中的网民自我意识研究[D]: [博士学位论文]. 北京: 中国传媒大学, 2008.
[11] 王娟, 兰月新, 张鹏. 面向突发事件的意见领袖对网民情感作用机理及实证研究[J]. 山西警察学院学报, 2023, 31(3): 45-53.
[12] 迟钰雪, 刘怡君. 基于社团发现的舆论意见领袖影响下“多对多”用户交互模式研究[J]. 电子科技大学学报(社科版), 2018, 20(3): 16-21.
[13] 朱志国, 张翠, 丁学君, 等. 基于熵权灰色关联模型的重大突发舆情意见领袖识别研究[J]. 情报学报, 2017, 36(7): 706-714.
[14] 张连峰, 周红磊, 王丹, 等. 基于超网络理论的微博舆情关键节点挖掘[J]. 情报学报, 2019, 38(12): 1286-1296.
[15] Xu, P.P., Wu, Y.C., Wei, E.X., et al. (2013) Visual Analysis of Topic Competition on Social Media. IEEE Transactions on Visualization and Computer Graphics, 19, 2012-2021. [Google Scholar] [CrossRef] [PubMed]
[16] 王晨, 谭皓然, 任雪娇. 意见领袖对网络用户“种草”行为的影响研究——以小红书为例[J]. 现代商业, 2025(1): 43-46.
[17] Grootendorst, M. (2022) BERTopic: Neural Topic Modeling with a Class-Based TF-IDF Procedure.
[18] Ma, J., Wang, L., Zhang, Y., Yuan, W. and Guo, W. (2023) An Integrated Latent Dirichlet Allocation and Word2vec Method for Generating the Topic Evolution of Mental Models from Global to Local. Expert Systems with Applications, 212, Article ID: 118695. [Google Scholar] [CrossRef
[19] 张教萌, 师荣蓉. 基于利益相关者的突发事件网络舆情演化研究[J]. 情报学报, 2022, 41(5): 525-535.
[20] 卢兴, 刘鹏宇. 失序与共治: 体育突发事件网络暴力生成路径与协同防控——基于体育突发事件网络舆情案例的实证研究[J]. 武汉体育学院学报, 2024, 58(5): 19-26.