Playlists are a natural delivery method for music recom- mendation and discovery systems. Recommender systems offering playlists must strive to make them relevant and enjoyable. In this paper we survey many current means of generating and evaluating playlists. We present a means of comparing playlists in a reduced dimensional space through the use of aggregated tag clouds and topic models. To evaluate the fitness of this measure, we perform prototypical retrieval tasks on playlists taken from radio station logs gathered from Radio Paradise and Yes.com, using tags from Last.fm with the result showing better than random performance when using the query playlist’s station as ground truth, while fail- ing to do so when using time of day as ground truth. We then discuss possible applications for this measurement technique as well as ways it might be improved.
Presented at the Workshop on Music Recommendation and Discovery on 26 September 2010, co-located with ACM Recommender Systems.