A Mood-Based Music Classification
and Exploration System

Mood classification of music is an emerging domain of music information retrieval. In the approach presented here features extracted from an audio file are used in combination with the affective value of song lyrics to map a song onto a psychologically based emotion space. The motivation behind this system is the lack of intuitive and contextually aware playlist generation tools available to music listeners. The need for such tools is made obvious by the fact that digital music libraries are constantly expanding, thus making it increasingly difficult to recall a particular song in the library or to create a playlist for a specific event. By combining audio content information with context-aware data, such as song lyrics, this system allows the listener to automatically generate a playlist to suit their current activity or mood.

Context-Aware Playlist Generation

Digital music libraries are constantly expanding, thus making it increasingly difficult to recall a particular song in the library, let alone create a playlist for a specific event. By using context-aware data the user can automatically generate a playlist to suit their current activity or mood. Songs corresponding to a certain tempo can be selected from the user's music library by converting their rate of step to a tempo value. Alternatively, a musical playlist can be generated based on the affective value of text entered by the user. Thus, music can be tailored to fit a specific context by using alternative classification methods and by monitoring a user's actions.

$100 Laptop (OLPC)

I have been developing several Csound/Python games and media players for the One Laptop Per Child initiative.