The Music Genome Project is more than a recommendation engine. It leverages big data to suggest music that listeners would be most-likely to enjoy based on several factors. These include a less-weighted “skip song” vs. a “thumbs down”, time of day, and more.
The Music Genome Project is primarily used by Pandora, which was started as result of the project, after it was met with mild industry response in 2005. It’s big data comes not just from its users and their interactions with the content, but also from actual musicians who rate songs on up to 450 possible characteristics to define a song.
Read more about how big data is being used by Pandora and the Music genome project here (http://www.forbes.com/sites/amadoudiallo/2013/10/06/pandora-radios-dominance-built-on-big-data-edge/).