Research On Book Recommendation System Based On Collaborative Filtering Technology

With the development of data bases and internet communications technology, most libraries in higher education institutions have transformed from the manual service mode to the computer-assisted service, especially in the field of book borrowing and returning and book cataloging. With the long period of application, the library bibliographic retrieval system has accumulated a large amount of information; it would improve the management and service if we can find out the collaborative rules behind that information.Data mining technology is vital in the application of library bibliographic retrieval system, it can keep the track of the readers’interests and trace the book borrowing rules, hence to recommend books to readers with relevant topics and make integrations of all the library resources. Using data mining technology effectively, we can better understand the needs of readers and improve the service, as well as upgrade the book management.Association rules and collaborative filtering are the two vital technologies in the field of data mining research. Association rules is to find out the relations among items in a large amount of information, collaborative rules can realize recommendation function by collaborating users with similar interests or items with similar attributes. This paper thoroughly explains how to proceed the data mining with the record of book borrowing and returning by using the association rules and collaborative filtering rules. Based on the authentic record of book borrowing and retuning, this paper manages to design a book recommendation system with reference to the user-based collaborative filtering technology and realize the personalized recommendation service. In the final part of this paper, it examines the validity and predictability of the system through certain experiments and proves to be effective for library recommendation service.