Using song social tags and topic models to describe and compare playlists

Ben Fields, Christophe Rhodes, Mark D'Inverno

Research output: Chapter in book/report/conference proceedingConference proceedingpeer-review

1 Citation (Scopus)

Abstract

Playlists are a natural delivery method for music recommendation 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 failing 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.

Original languageEnglish
Title of host publicationProceedings of the Workshop on Music Recommendation and Discovery 2010, WOMRAD-2010
PublisherCEUR-WS
Pages5-12
Number of pages8
Publication statusPublished - 26 Sept 2010
Event1st Workshop on Music Recommendation and Discovery 2010, WOMRAD 2010 - Barcelona, Spain
Duration: 26 Sept 201026 Sept 2010

Publication series

NameCEUR Workshop Proceedings
PublisherCEUR-WS
Volume633
ISSN (Print)1613-0073

Conference

Conference1st Workshop on Music Recommendation and Discovery 2010, WOMRAD 2010
Country/TerritorySpain
CityBarcelona
Period26/09/1026/09/10

User-Defined Keywords

  • Information retrieval
  • LDA
  • Metric space
  • Music
  • Playlists
  • Similarity
  • Social tags
  • Topic Models

Fingerprint

Dive into the research topics of 'Using song social tags and topic models to describe and compare playlists'. Together they form a unique fingerprint.

Cite this