Tuesday, September 11, 2012

Album Release Party: Network Visualization Edition

Last week, indie pop group The xx released their highly anticipated second album titled coexist. While the album has been available in stores, the band also streamed the album for free on their website — with a twist.

Every time a fan shared the streaming album via Facebook, Twitter, or email, the band tracked the geographic locations of the sharer and new listener. After compiling the data, they created an interactive data visualization tool detailing the album's spread around the globe. While listening to the entire album, you can watch as golden streams dance across the map, showing which regions have been infected with the band's beats and rhythms.

So where does physics come into play? Physics, especially statistical mechanics, has played a strong role in shaping the field of network theory that underlies this album release experiment. But maybe researchers could learn a thing or two from musicians like The xx. Experiments like this combined with social music players such as spotify and last.fm could reveal where the world's most influential tastemakers reside. Just add a pinch of motivated network researchers, and mix thoroughly.

A screengrab from The xx's data visualization tool.

Network researchers have drawn inspiration from physics for a number of diverse applications. On the Buzz blog, we've covered how scientists used Twitter to correctly predict the American Idol winner and simple models used to explain ideological conflict. For another article, I learned that triangulation used in cellular networks inspired researchers working to track down the source of an epidemic.

Tracking the spread of music seems like a straightforward application of earlier research on social networks. The data are certainly there. A decade ago, Last.fm pioneered audio scrobbling — a way to track what users are listening to and give them relevant recommendations. Since then, billions of songs have been scrobbled on the site. More recently, social music player Spotify has surged in popularity, partly due to its integration with Facebook.

Much of this data on what people are listening to is also geo-tagged, which would allow researchers to track the spread of a new album or song. I'd certainly be interested in seeing how certain bands' popularity spreads geographically.

More broadly, researchers could couple this data with musical recommendations like the ones created by last.fm. Perhaps research could eventually reveal emerging trends or genres in music — even before Pitchfork identifies them.

After a cursory search online, there seems to be little interest among researchers in tracking the spread of musical tastes. Nonetheless, the advent of huge listening datasets and advances in network theory could lead to some interesting findings. Hopefully, physicists and computer scientists are listening.

You can play around with The xx's data visualization tool on their website, or click on the embedded box below. You can control the music on the bottom left hand screen, and the top right sliding scale enables you to track the album's spread over the last week.


If you want to keep up with Hyperspace, AKA Brian, you can follow him on Twitter.

No comments:

Post a Comment