Publication

Abstract : Recommender systems are becoming increasingly ubiquitous, powering the way we discover new products, services, and information. However, the black-box nature of these systems can make it difficult for users to understand why they are being recommended certain items. So, transparency, trust, scrutability, and fairness in the recommendations generated by the recommender system are crucial for the users to consider. The tools to make the recommendations trustable, explainable, and scrutable are the transparent recommender systems. Transparent recommender systems aim to address these problems by providing users with more information about how their recommendations are generated. This information can include the factors that were considered, the relative importance of those factors, and the reasoning behind the final recommendation. This chapter provides a detailed study of transparent AI, its benefits and challenges, and open-source packages available to make the machine learning models understandable. It provides an outline of transparent recommender systems and includes the most recent work in this domain.