In this paper, we address propose a user-aware music auto-tagging system. Specifically, we use both the audio content and user information extracted from the user listening history to predict contextual tags for a given user/track pair. We also propose a new dataset of music tracks annotated with contextual tags per user.
Karim Ibrahim, Elena Epure, Geoffroy Peeters, Gael Richard. SHOULD WE CONSIDER THE USERS IN CONTEXTUAL MUSIC AUTO-TAGGING MODELS?. 21st International Society for Music Information Retrieval Conference, Oct 2020, Montreal, Canada.