Application of Data Mining in the Field of University Libraries for Book Borrowing Services

Abstract

University libraries are knowledge resource repositories open to all teachers and students. Using data mining techniques to explore the hidden relationships between readers and books in the borrowing data can maximize the utilization of information and better apply it to the library field. This article uses cluster analysis to implement personalized recommendation algorithms for books. Two recommendation algorithms based on book clustering and reader clustering were proposed, and a hybrid recommendation algorithm was formed by combining the two algorithms. A case analysis was conducted. The results showed that the hybrid recommendation algorithm provided more personalized and specific recommendations and had a higher accuracy than the other two algorithms regardless of how many books the target user borrowed.