[目的/意义] 当前学术交流行为已拓展到线上多场景互动，知乎、小红书等内容社交平台凭借庞大的用户基数与海量话题，正成为学术交流的“新场域”。研究内容社交平台上的学术交流主题特征，有助于优化平台学术交流功能、提升用户学术信息素养，促进更多用户开展高质量学术交流，进而加速科研创新与知识传播。 [方法/过程] 首先，基于S-O-R理论构建内容社交平台学术交流行为研究框架；其次，利用python爬虫技术采集2020年1月—2025年10月知乎相关学术话题及小红书学术类标签下的互动评论文本，经筛选、去重、同义词合并等预处理后，运用jieba分词与停用词过滤进行文本处理，并基于大连理工大学情感词汇本体库开展情感分类；最后，采用BERTopic主题模型挖掘主题特征，并对内容社交平台的学术交流行为进行深入分析。 [结果/结论] 已有大量用户选择在社交平台评论区开展学术交流与互动。基于S-O-R理论构建的用户行为特征主题模型可有效揭示用户在非正式学术交流中的核心需求。研究结果显示，学术类主题主要集中在文献阅读与科研实践、学术心理与情感共鸣、学术社交与职业发展、学科领域与机构交流等方面。上述结论可为社交平台优化学术交流功能、提升交流质量、引导健康学术生态提供实证依据。

[Purpose/Significance] Academic communication has now expanded into multi-scenario online interaction. Content-based social platforms such as Zhihu and Rednote, with their massive user bases and abundant topics, are becoming new venues for academic exchange. Investigating the thematic characteristics of academic communication on content-based social platforms can help optimize the platforms' academic communication functions, enhance users' academic information literacy, and encourage more users to engage in high-quality academic exchange, thereby accelerating research innovation and knowledge dissemination. [Method/Process] First, a research framework for academic communication behavior on content-based social platforms was constructed based on the S-O-R theory. Second, Python web-crawling techniques were employed to collect interactive comment texts under Zhihu's academic topics and Rednote's academic tags from January 2020 to October 2025. After preprocessing including screening, deduplication, and synonym merging, jieba word segmentation and stop-word filtering were applied for text processing, and sentiment classification was conducted based on the Dalian University of Technology sentiment lexicon. Finally, the BERTopic topic model was adopted to mine thematic characteristics, and an in-depth analysis of academic communication behavior on content-based social platforms was carried out. [Result/Conclusion] A large number of users have already chosen to conduct academic communication and interaction in the comment sections of social platforms. The user behavior characteristic topic model constructed on the basis of the S-O-R theory can effectively reveal users' core needs in informal academic communication. The findings show that academic topics are mainly concentrated in such aspects as literature reading and research practice, academic psychology and emotional resonance, academic social networking and career development, and disciplinary fields and institutional exchange. These conclusions provide empirical evidence for social platforms to optimize academic communication functions, improve the quality of communication, and guide a healthy academic ecosystem.