Workshop on Music Recommendation and Discovery

In the last decade, digital music has transformed the landscape of music experience and distribution. Personal music collections can exceed thousands of tracks, while the Internet has made it simpler than ever to find and access music. In this scenario, music recommendation systems have become increasingly important for listeners to discover and navigate music.

Music-centric recommenders such as Last.fm and Pandora have enjoyed commercial and critical success. But how well do these systems work? How good are the recommendations? How far into the “long tail” can they go before surrendering to bad quality works?

The approach of recommending songs as if they were books is limiting; better results can be achieved by taking into account the peculiarities of music and the music recommendation process. A successful music recommender should combine insights from user preferences (classical collaborative filtering) with the content (audio analysis, tags, lyrics, etc..) while integrating the social interactions along with the psychological and emotional aspects connected to music consumption.

The Workshop on Music Recommendation and Discovery was meant to be a platform where the Recommender System, Music Information Retrieval, User Modeling, Music Cognition, and Music Psychology communities can meet, exchange ideas and collaborate.

  • Amélie Anglade
  • Claudio Baccigalupo
  • Norman Casagrande
  • Òscar Celma
  • Paul Lamere
Workshop Date

September 26, 2010

Web site