Automatic playlist continuation using subprofile-aware diversification

Research output: Chapter in Book/Report/Conference proceedingsChapterpeer-review

Abstract

The ACM RecSys Challenge 2018 involves the task of automatic playlist continuation (APC), aiming to help users to create and extend their own music playlists. In this paper, we explain teamrozik's approach to the Challenge. Our approach to APC is twofold: Cold-Start-APC for short playlists and SPAD-APC for other playlists. Cold-Start-APC is a rudimentary popularity-based recommender. SPAD-APC treats playlists as if they were user profles. It builds an implicit matrix factorization model to generate initial recommendations. But it re-ranks those recommendations using Sub Profle-Aware Diversification (SPAD), which is a personalized intent-aware diversification method. The SPAD re-ranking method aims to ensure that the final set of recommendations covers different interests or tastes in the playlists of the users, which we refer to as subprofles. We show that such subprofles do exist within playlists and we show that the SPAD method achieves higher precision than matrix factorization alone.

Original languageEnglish
Title of host publicationProceedings of the ACM Recommender Systems Challenge 2018, RecSys Challenge 2018
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450365864
DOIs
Publication statusPublished - 2 Oct 2018
Event12th ACM Recommender Systems Challenge Workshop, RecSys Challenge 2018 - Vancouver, Canada
Duration: 2 Oct 20182 Oct 2018

Publication series

NameACM International Conference Proceeding Series

Conference

Conference12th ACM Recommender Systems Challenge Workshop, RecSys Challenge 2018
Country/TerritoryCanada
CityVancouver
Period2/10/182/10/18

Keywords

  • Automatic playlist continuation
  • Diversity
  • Music recommender
  • Playlists
  • Subprofles

Fingerprint

Dive into the research topics of 'Automatic playlist continuation using subprofile-aware diversification'. Together they form a unique fingerprint.

Cite this