Tagged at first listen: an examination of social tagging practices in a music recommender system
| Autor | Audrey Laplante |
| Cargo | Associate professor, École de bibliothéconomie et des sciences de l'information, Université de Montréal |
| Páginas | 33-54 |
33
v. 20, n.esp, 2015
p. 33-54
ISSN 1518-2924
Encontros Bibli: revista eletrônica de bibliotecono mia e ciência da informação, v. 20, n. esp. 1, p.
33-54 Fev., 2015. ISSN 1518-2924. DOI: 10.5007/1518-2924.2015v20nesp1 p33
Tagged at first listen: An examination of social tagging
practices in a music recommender system
Etiquetagem em primeira audição: uma análise das
práticas de etiquetagem social em um sistema de
recomendação de música
Audrey LAPLANTE1
Abstract
Social tagging has become a very common way to index different types of resources on the
web. Less prevalent in music than in other domains, social tagging is nevertheless used in a
popular recommender system, Last.fm. Although the number of publications on tagging and
folksonomies has exploded in the last few years, music tagging is still not well studied. In this
paper, we present a study of tagging practices of Last.fm users. W e examine the social tagging
of songs during the first th ree months after their release. Our analysis show s that the release
of a song triggers a burst in tagging activity that lasts two weeks, after what it decreases
sharply and then remain s fairly constant for the next ten weeks. We also find that a majority
of songs do not get tagg ed during the first week and that tagging was positively related to
popularity. Final ly, we f ind that tags that have been frequently applied to a given song are
more likely to be genre related, shorter in length, and relatively objective than tags that have
been applied only once.
Keywords: Social tagging; music indexing; music recommender systems
Resumo
A etiqu etagem social (social tagging) tornou-se uma forma muito comum de index ar
diferentes tipos de recursos na web. Menos predominante na música do que em outros
domínios, a etiquetagem social é utilizada em um popular sistema de recomendação, Last.fm.
Embora o número de publicações sobr e atribuição de tags (etiquetas) e f olksonomia tenha
explodido nos últimos anos, a atribuição de tags à música permanece pouco estudada. Neste
artigo, apresentamos um estudo das práticas d e atribuição de tags dos usuários do L ast.fm.
Examinamos a etiquetagem social de músicas durante os primeiros três meses após sua
liberação. Nossa análise mostra que a libera ção de uma música desencadeia um a explosão
nas atividades de atribuição de tags que dura duas semanas, depois essa atividade diminui
de forma acentuada e, então, permanece razoavelmente constante nas próximas 10 semanas.
Também verificamos que a maioria d as músicas não recebem tags durante a primeira
semana e que a atribuição de tags foi positivamente re lacionada à popularidade. Finalmente,
constatamos que as tags que são frequentem ente aplicadas a determinada música são mais
relacionadas ao gênero, são menores em extensão e r elativamente mais objetivas do que tags
que tenham sido aplicadas uma única vez.
Palavras-chave: Etiquetagem social; tag; indexa ção de música; sistemas de recomendação
de música
Esta obra está licenciada sob uma Licença Creative Commons.
1 Associate professor, École de bibliothéconomie et des sciences de l’information, Uni versité
de Montréal - audrey.laplante@umontreal.ca
ARTIGO
Recebido em:
20/05/2014
Aceito em:
30/11/2014
34
1. INTRODUCTION
Indexing music for retrieval and discovery has become a much more complicated
task in the last decade. Although indexing music albums with two or three subject
headings in library catalogues or classifying albums by broad genres in music stores is
often sufficient to allow users to find what they are looking for, these traditional indexing
and classification methods cannot be maintained in current digital music libraries. Indeed,
whereas brick-and-mortar music stores typically keep a few thousand unique albums on
their shelves, online music retailers and on-demand streaming services give access to
millions of tracks (for example, the music streaming service Spotify1 has more than 20
million songs in its catalogue2 and the online music retaileriTunes Store, more than 37
million songs3). Therefore, broad genres become almost useless since they are not
discriminative enough. For instance, a search with “alternative” in Amazon MP3retrieves
over2 million songs, hence the need to index music with more specific terms, and at the
song level rather than at the album level.
To reach that level of indexing and be able to provide personalized music
recommendations to its users, Pandora4has taken a rather traditional approach and
relies on music analysts to index songs, with up to 450 attributes per song5. This
approach is obviously time consuming and costly, which translates into a lower
growth rate: Pandora’s catalogue currently contains just over one million tracks6,
which is well below most music services. By comparison, another popular
recommender service, Last.fm7, opted for a more scalable and affordable model. It
relies on its users for indexing. Last.fm’s users are encouraged to assign free-form
textual labels (i.e., social tags) to songs, albums, and artists. Although Last.fm does
not make public the number of songs in its catalogue, it can be estimated to several
millions8.
The terms “social tagging” and “collaborative tagging” refer to the process by
which people assign tags or keywords, whereas the sum of all tags applied by the
1http://www.spotify.com
2http://press.spotify.com/us/information/
3http://www.apple.com/itunes/features/#store
4http://www.pandora.com
5http://www.pandora.com/about/mgp
6http://blog.pandora.com/press/pandora-company-overview.html
7http://www.last.fm
8Last.fm does not reveal the exact number of tracks in its catalogue. However, accordi ng to CBS
(the owner of Last.fm), ithad more than 12 million tracks i n 2013 (the page was taken down in the
Fall of 2013 and can be viewed on Internet Archive at
https://web.archive.org/web/20130913010821/http://cbsimg.com/our-products/last-f m).
Para continuar a ler
PEÇA SUA AVALIAÇÃO