As one of the tools to provide accurate services in the era of big data, user profile has attracted widespread attention. In order to improve the personalized service ability of university library, the construction of user portrait model is imperative. Based on the data of university library, this paper uses k-means algorithm to analyze the user profile. Finally, taking the relevant data of Fudan University Library as an example, this paper analyzes each cluster of user profile, and finally puts forward corresponding suggestions for the development of the library according to the results of user profile.