|Presentation Date||March 30, 2013|
|Topic(s)||Music Genome Project|
Automatic music recommendation systems require large databases of music data and complex rules to provide the user with decent recommendations. One such source of data comes form the Music Genome Project. This project aims to "capture the essence of music at the fundamental level" using hundreds of attributes to describe and store individual songs. Pandora Radio, is a well-known example of a music recommendation service, in which "radio stations" are created to fit the user's unique tastes in music. Pandora Radio uses and owns the Music Genome Project.
I want to write a paper that looks at how the Music Genome Project is used within Pandora and compare the Pandora Radio service to other known music recommendation systems. The other systems include: MyStrands, iTunes Genius, Spotify, and Last.fm. It would be interesting to compare the systems' respective databases and algorithms used to create their recommendations.
References: How Pandora Radio Works, http://computer.howstuffworks.com/internet/basics/pandora.html MashTable, http://mashable.com/2009/02/02/music-recommendation-services/ Music Genome Project, http://en.wikipedia.org/wiki/Music_Genome_Project Spotify Radio, http://www.digitaltrends.com/music/echo-nest-the-secret-weapon-behind-the-new-spotify-radio/ Last.fm, www.last.fm Four Approaches to Music Recommendation, http://readwrite.com/2009/01/26/music_recommendations_four_approaches