Recommender System for Books in University Library with Implicit Data

Recommender system is a very important tool to help customers make choices more easily in a large variety of offered products. However, it is difficult to make directly use of the recommender system to provide suggestion for the traditional books in a library because of the shortage of the explicit feedback, like readers’ rating, reviews etc. We propose a model that transfers the implicit data of readers borrow history to explicit data and apply the SVD++ algorithm in the recommender system.