Abstract: [Purpose/Significance] This study aims to quantitatively analyze the themes of user-generated content on the Xiaohongshu platform, with the goal of improving the content quality of social media platforms and promoting their healthy development. [Design/Methodology] Based upon the Censydiam motivation analysis model, this study utilizes the BERTopic topic extraction method to construct a framework for researching the themes and motivations behind information sharing in online communities. The study focuses on the Xiaohongshu platform, collecting 3,688,942 pieces of user-generated content to analyze the users' motivations behind information sharing across different themes. [Findings/Conclusion] The themes of information sharing by Xiaohongshu users primarily encompass life experiences, shopping, and food. By precisely pinpointing these themes using the Censydiam model, the study reveals that users' primary needs include a sense of belonging and identification with social groups, social integration and communication, as well as seeking enjoyment and relaxation. Therefore, the platform should place greater emphasis on maintaining alignment with the overall social media environment and ensuring a consistent social tone to reflect care for the user community and enhance their sense of belonging. [Originality/Value] This study is based on textual data from the Xiaohongshu online community, verifying the feasibility of the research framework and offering new insights and methods for the analysis and development of online communities.

Key words: Topic extraction, Information sharing, Motivation analysis, Censydiam model, Xiaohongshu