Recommending playlists to users in the context of a digital music service is a difficult task because a playlist is often more than the mere sum of its parts. We present a novel method for generating playlist embeddings that are invariant to playlist length and sensitive to local and global track ordering. The embeddings also capture information about playlist sequencing, and are enriched with side information about the playlist user. We show that these embeddings are useful for generating next-best playlist recommendations, and that side information can be used for the cold start problem.
Matrix Factorization (MF) is a common method for generating recommendations, where the proximity of entities like users or items in the embedded space indicates their similarity to one another. Though almost all applications implicitly use a Euclidean embedding space to represent two entity types, recent work has suggested that a hyperbolic Poincar\'e ball may be more well suited to representing multiple entity types, and in particular, hierarchies. We describe a novel method to embed a hierarchy of related music entities in hyperbolic space. We also describe how a parametric empirical Bayes approach can be used to estimate link reliability between entities in the hierarchy. Applying these methods together to build personalized playlists for users in a digital music service yielded a large and statistically significant increase in performance during an A/B test, as compared to the Euclidean model.