Book Recommendation System using Data Mining for the University of Hong Kong Libraries

Abstract

This paper describes the theoretical design of a Library Recommendation System, employing k- means clustering Data Mining algorithm, with subject headings of borrowed items as the basis for generating pertinent recommendations. Sample data from the University of Hong Kong Libraries (HKUL) has been used in a Quantitative approach to study the existing Library Information System, Innopac. Data Warehousing and Data Mining (k-means clustering) techniques are discussed. The primary benefit of the system is higher quality of academic research ensuing from better search results. Personalization improves individual effectiveness of learners and overall in better utilizing library resources.