What music do people listen to? How does their taste change with time? Where do new music styles come from?
A.Track.Tion is a data visualization aimed at shedding light on these deep and interesting questions. The core dataset is the Whitburn Project list, a collective effort to gather historical weekly music sales rankings published by the Billboard company (the project is now maintained at the Bullfrogs Pond. The list goes back up to 1890, but data is more complete after 1954. The subset we used (after cleaning of the original) contains 33560 songs. You can see (and play with) random songs from the database in the Serendipity section.
In order to explore the popularity of music, we had to expand the data in two ways: We collected genre information of the songs from Wikipedia, and we computed an estimate of the percentage of total sales associated with a given ranking position held for a week (see here for methodology details).
We coupled our new datasets with estimates of global music sales, resulting in our final visualization of popularity of music genres and subgenres. The plot depicts broad observations such as the dominance of rock throughout time, and also smaller stories such as the first and second rise of hip hop, or the quick burst of disco (a sub-genre of electronic music) in the 70's.
The genres extracted from Wikipedia contained information about which styles influenced which others. Our naive expectation of a "tree of music" styles was quickly shattered, resulting in a complex and organic-like tangled mess of relationships (again, see our methodology for more details and a discussion of the limitations of these data). Influence allows us to explore and discover unexpected relationships (such as Cowpunk, the mixture of Country and Punk music).
By coupling this graph to our popularity calculations we can also see that the most influential genre (Electronic music) is not (by far) the most popular one. In other words, Popularity is what people listen to, while Influence is what musicians listen to.
This project was financed by the Barcelona Supercomputing Center and developed by the Scientific Visualization Group -- Fernando Cucchietti, Diana Fernanda Vélez, Luz Calvo and Guillermo Marín.