Design of Library User Profile System Based on Dynamic Density Clustering Algorithm and Stream Computing
Jingpei Liao · 2021
The study of book recommendation system based on user profile is of great significance in accurately grasping the potential reading needs of readers, improving the quality of book recommendation and promoting the development of personalized service in libraries, and it is an indispensable part of the future construction of smart libraries. The existing book recommendation system based on user profile has a series of problems, such as incomplete summary of various information of users, inaccurate grasp of users' real reading needs, and time lag in user information analysis, etc. This paper describes the architecture principle, label system and algorithm flow of library user profile system, and uses dynamic density clustering method and stream computing based on time series analysis to give the label system a time dimension by combining the big data characteristics of users. Combined with the needs of the development of high-quality personalized information services in future smart libraries, this paper explores the application of this system in the effective study of library reading promotion activities, readers' sharing sessions, readers' personalized reading list design and so on.