blog.shrub

there are a bunch of ways to analyze and recommend music.

• shazam, which uses acoustic fingerprinting to recognize songs based on 15-second excerpts
• transpose’s goombah and the indy project, which use collaborative filtering to recommend music
• music plasma, an interactive flash application which illustrates (perceived) relationships between artists.
• hit song science, which predicts where or not a song will be a hit.

recently, the Wall Street Journal featured the music genome project, which paid a bunch of musicians to listen to thousands of songs and evaluate them on a bunch of objective criteria.

This entry was posted on Tuesday, October 18th, 2005 at 12:20 am and is filed under music. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.