[目的/意义]挖掘用户数据并构建在线健康社区用户画像，有利于深入了解用户需求，提高在线健康社区的用户使用体验，促进在线健康社区发展。[方法/过程]结合在线健康社区建设现状，使用RFM模型筛选典型用户，并从事实维度、模型维度和预测维度构建在线健康社区用户画像标签，以问卷调查数据为依据，通过K-MEANS聚类分析实现部分用户画像实证研究。[结果/结论]数据驱动背景下构建在线健康社区用户画像可有效实现个性化检索和精准推送，有利于增强用户黏性，助力网站推广，对提升在线健康社区平台精准服务水平具有重要的参考价值。

[Purpose/significance]Mining user data and constructing online health user portraits is conducive to in-depth understanding of user needs,improving user experience in online health communities,and promoting the development of online health communities.[Method/process]Combining with the current status of online health community construction,use the RFM model to screen typical users and construct online health community user portrait tags from fact,model,and prediction dimensions.Based on the questionnaire data,use K-MEANS cluster analysis to achieve the part empirical analysis of portraits.[Result/conclusion]Constructing online health community user portraits in a data-driven context can effectively achieve personalized retrieval and accurate push,which is conducive to enhancing user stickiness and assisting website promotion,and has important reference value for online health community platforms to improve accurate service levels.