Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper addresses the item cold-start problem in music recommendation by exploiting the collaborative signal available at the artist level. They introduce ACARec, an attention-based architecture that generates collaborative filtering embeddings for new tracks by attending over the artist's existing catalog. Experiments demonstrate that ACARec more than doubles Recall and NDCG compared to content-only baselines, particularly improving new artist discovery and cold item popularity estimation.
New music recommendation doesn't have to be a shot in the dark: tapping into an artist's existing catalog can more than double recommendation accuracy for their new tracks.
The item cold-start problem poses a fundamental challenge for music recommendation: newly added tracks lack the interaction history that collaborative filtering (CF) requires. Existing approaches often address this problem by learning mappings from content features such as audio, text, and metadata to the CF latent space. However, previous works either omit artist information or treat it as just another input modality, missing the fundamental hierarchy of artists and items. Since most new tracks come from artists with previous history available, we frame cold-start track recommendation as 'semi-cold' by leveraging the rich collaborative signal that exists at the artist level. We show that artist-aware methods can more than double Recall and NDCG compared to content-only baselines, and propose ACARec, an attention-based architecture that generates CF embeddings for new tracks by attending over the artist's existing catalog. We show that our approach has notable advantages in predicting user preferences for new tracks, especially for new artist discovery and more accurate estimation of cold item popularity.