Fruitr is an experiment with the Echonest API.
It is an exposition of the idea stated here: http://www.rishighan.com/thesis/
The initial idea calculates the Pearson product-moment coefficient of correlation between the hotttnesss and familiarity ratings retrieved from Echonest. This helps to add a layer of verification on top of the numerical ratings that Echonest applies to intangible concepts such as how 'hot' an artist is.
As I build upon the idea, I am also creating a Ruby wrapper around most of the Echonest API hooks. If you would like to contribute, you can send a pull request.