Policy by the Numbers: Music metadata in the Commons

Music metadata in the Commons

Wednesday, January 11, 2012

In 2010, I began a year-long mixed-methods study of MusicBrainz, a community music metadatabase with companion open source software that cleans up the metadata on digital music files. Studies have been conducted on various aspects of commons-based, peer-produced projects, notably free and open source software and Wikipedia. But MusicBrainz is unique: MusicBrainz contributors play the role of information scientists for this data commons, working as digital librarians, standards-setters and catalogers of music.
Understanding what drives people to voluntarily curate and contribute to a data commons benefits our overall understanding of how these commons work. If we find common characteristics among a few successful data communities, we can inform the design of data commons for other domains so that they are more likely to thrive. What follows is a snapshot of some of the more interesting findings. Full report with methodology is here, and a quick presentation deck is available here.

The importance of open source
MusicBrainz editors believe that information resources and music metadata should be free (see table). In interviews, they discussed openness as a “philosophy,” and knowing that they are building a bigger and better project useful to others works as an intrinsic motivation.

One editor describes the process of using and adding data as a “virtuous circle.” He contributes to a variety of peer-produced projects, including Wikipedia, and called himself “selfish” when asked why he contributes to open source projects. He said, “I think people who really value things will want to ensure they continue. And there are two ways you can do it. One, you can use your wallet. The other one is, if it's an option, you can contribute and make it a better thing.” Another editor echoed that sentiment: “I get to benefit from MusicBrainz and this is somewhat of my payback to the community.”

Discovery through Contribution
Several of the editors I spoke with told me stories about having discovered an artist or a collaboration in the process of interacting with the data, whether by browsing or editing data. One of my hypotheses was that because of the patterns of exposure, editors who have discovered an artist through MusicBrainz are likely to have entered on average more edits—spending more time with the database—than those who have not.

To test this, I compared the means of those who answered “Yes” and those who answered “No” against the log of the number of edits entered. Results, shown above, support the hypothesis that those who have discovered an artist through MusicBrainz have made more edits on average than those who have not. Engaging with the data benefits contributors beyond just providing accurate music metadata.

Where are the women?

My sample was overwhelmingly male, and follow-up interviews indicated that most of the editors active in communication channels are male. This gender disparity is not unique to MusicBrainz. This NetworkWorld article cites several sources that showed lower female participation in F/OSS projects and in Wikipedia. This is definitely an area to explore further.

by Jess Hemerly, Senior Policy Analyst at Google

Policy by the Numbers

Music metadata in the Commons

No comments :

Labels

Archive

Feed

Disclaimer

Company-wide

Products

Developers