不同社交媒体的舆情比较研究：以“胖猫事件”为例

A Comparative Study of Public Opinion on Different Social Media Platforms: The Case of the “Fat Cat Incident”

摘要: 研究目的：互联网加速了舆情事件在社交媒体平台上的传播，除了在以文本内容为主的微博上快速传播外，视频网站也是舆情发酵的重要平台。既有舆情研究多聚焦微博，对视频网站的研究较少，更缺乏对两者的对比。以“胖猫事件”为案例，对比研究微博和Bilibili用户的话题和态度。研究方法：利用TF-IDF提取文本特征词及权重，结合LDA主题聚类识别焦点词和核心议题；使用百度情感分析API，比较不同平台和主题下的用户情感。研究结果：微博和Bilibili在主题和情感表现上既有相似性，也存在差异：1) 主题内容：微博和Bilibili前期的讨论主要集中在情感与性别话题；但警方通报后，微博侧重对真相的反应和网络舆论的讨论，Bilibili则延续了前期对情感和家庭影响的关注。微博讨论主题相对分散，而B站较为集中。2) 情感表现：负面情感在两平台均占主导，但微博更为强烈。这与微博的快速传播和部分蹭热度行为密切相关，微博上“流量与热度”和“网络舆论”两个主题的负面情绪尤其高。综上，两平台舆情主题和情绪有一定相似性，但微博蹭热度现象更明显，这与文本为主的社交平台信息发布门槛低有关，而视频网站的讨论相对更集中更深入。这一发现为社交媒体管理和舆情监控提供了有价值的参考。

Abstract: Research purpose: The Internet has accelerated the dissemination of public opinion events on social media platforms. In addition to the rapid dissemination on text-based platforms like Weibo, video websites are also important platforms for the escalation of such events. Existing research on public opinion predominantly focuses on Weibo, with comparatively little attention given to video platforms, and even fewer studies offering a comparative analysis between the two. Using the “Fat Cat Incident” as a case study, this research conducts a comparative analysis of user topics and attitudes on Weibo and Bilibili. Research method: Using TF-IDF to extract textual feature words and weights, combined with LDA topic clustering to identify focus words and core topics; using Baidu Sentiment Analysis API to compare user sentiment under different platforms and topics. Research result: Weibo and Bilibili exhibit both similarities and differences in thematic focus and emotional expression: 1) In the early stages, discussions on both platforms centered primarily around emotions and gender-related issues. However, after the police report was released, Weibo shifted its focus to reactions to the truth and broader public opinion, whereas Bilibili continued to emphasize the emotional and familial impacts. Discussions on Weibo were more dispersed, while Bilibili’s were more concentrated. 2) Emotional Expression: Negative emotions dominated on both platforms, but they were more intense on Weibo. This heightened negativity is closely linked to Weibo’s rapid dissemination and instances of opportunistic content for attention-seeking. In particular, negative sentiment was significantly higher in topics related to “traffic and trending” and “online public opinion” on Weibo. In summary, while there are notable similarities in the themes and emotions across both platforms, the phenomenon of attention-seeking is more pronounced on Weibo, likely due to the lower barriers to content dissemination on text-based social media. In contrast, discussions on video platforms like Bilibili tend to be more focused and in-depth. These findings offer valuable insights for social media management and public opinion monitoring.