Connecting the Dots: Reader Ratings, Bibliographic Data, and Machine-Learning Algorithms for Monograph Selection
Jingshan Xiao, Wenli Gao
Abstract
Recommender systems, a subclass of information filtering designed to predict the rating or preference of a user, are among the most successful examples of machine learning in action. Drawing inspiration from the benefits of using recommender systems for business, and their success in heightening the perceived utility of recommendations, this project was developed using Python to optimize collection recommendations and to help librarians make collection decisions using a recommender system. This paper illustrates several examples of building recommender systems using a variety of recommendation techniques to aid in the selection of monographs. It also points out possible future uses of recommender systems in libraries.