Abstract: 【Purpose/significance】Through the collection of users' online comments on the public culture cloud platform, this paper discusses the types of users' demands for the operation of the public culture cloud platform, constantly optimizes the ways and content forms of public cultural activities according to the characteristics of the demands, and improves the service quality of the public cul⁃ ture cloud platform.【Method/process】Collect online messages and online comments from users of public culture cloud platform by web crawler technology, extract keywords in comments by Keybert algorithm and carry out text coding, and finally form 8 user demand elements. According to these demand elements, Kano questionnaire is designed and investigated. Mixed class analysis, Kano model classification and Better-Worse coefficient analysis are carried out respectively to define the user's demand type.【Result/conclusion】 The research found that points reward, cultural characteristics and publicity and promotion are attractive needs, website operation and maintenance, resource acquisition are basic needs, community construction is expected needs, website booking efficiency and activity forms are undifferentiated needs. According to different demand types, relevant suggestions are put forward for the improvement direc⁃ tion of the public culture cloud platform to support further improvement and optimization of the public culture cloud platform【. Innova⁃ tion/limitation】The analysis of the user needs of the public culture cloud platform in this paper is conducive to improving the overall service level of the public culture cloud platform, but due to the incomplete data collection and limited sample size, its scientificity needs to be further investigated.