Diversity in Recommender Systems

Recommender systems are ubiquitous in our everyday life, providing us with recommendations for products in web shops, music, movies, or friends on social media. One of the goals of recommender systems is to provide users with a sufficiently diverse list of recommendations (for instance, covering different artists in a music recommendation list). In this thesis, we aim to analyze the performance of recommender algorithms w.r.t. diversity and the impact of optimizing for diversity on traditional rating prediction and ranking metrics.