Abstract: While providing services to users, university libraries generate a large amount of service data, including information on book borrowing, book reservation, and user access. Information of such can help managers analyze the behavior of users to further improve the quality of library services. This paper mainly studies the behavioral characteristics of group users and individual users, performs appropriate data analysis procedures and designs a reasonable visualization method for different dimensions, such as borrowing preferences and the site visit rules of different groups of users and individual users. For group users, static data analysis is mainly used, and for individual users, an interactive system is used for analysis. Finally, this paper applies the research method to real library data to prove its feasibility and effectiveness. 

Key words: Behavioral characteristics, Data analysis, Visualization, University library