User-aware music auto-tagging with contextual tags

<p>This is a user-aware music dataset labeled with the contextual use of each track according to each user. The dataset is composed of 10 contextual tags extracted based on user&#39;s usage through created playlists in the Deezer catalog. The tags are: &quot; car, gym, happy, night, relax, running, sad, summer, work, workout&quot;. For each track/user pair, a contextual tag&nbsp;is associated with it indicating that the user&nbsp;listens to the track in the associated context. Additionally, the users are represented as embeddings based on their listening history computed through the matrix factorization of the user/track matrix.</p> <p>The creation of the dataset and the&nbsp;baseline of our auto-tagging model is described in the paper: Karim M. Ibrahim,&nbsp;Elena V. Epure, Geoffroy Peeters,&nbsp;and Ga&euml;l Richard. &quot;Should we consider the users in contextual music auto-tagging models?&quot;&nbsp;<em>21st International Society for Music Information Retrieval Conference (ISMIR)</em>.&nbsp;2020. The source code of the paper is available here:&nbsp;<a href="https://github.com/KarimMibrahim/user-aware-music-autotagging">https://github.com/KarimMibrahim/user-aware-music-autotagging</a></p> <p>The dataset is composed of the SONG_ID&nbsp;which is the ID of the track in the Deezer catalog. Each track/user pair&nbsp;is labeled with each tag as either 1 (indicating a track&#39;s&nbsp;presence in the context) or 0 (indicating a track&#39;s absence). The 30 seconds track previews used to train the model in the paper can be accessed through the Deezer API:&nbsp;<a href="https://developers.deezer.com/api">https://developers.deezer.com/api</a>. Each user is represented with an anonymized USER_ID which is associated with the user embedding available in the user_embeddings.csv file.&nbsp;</p>