Adaptive jukebox : a context-sensitive playlist generator

Abstract

Nowadays, a lot of users own large collections of music MP3 files. Manually organising such collections into playlists is a tedious task. On the other hand random playlist generation may not always provide the user with an enjoyable experience. Automatic playlist generation is a relatively new field in computer science that address this issue, developing algorithms that can automatically create playlists to suit the user’s preferences. This paper presents our work in this field, where we suggest that playlist generators should be more context-sensitive. We also present Adaptive Jukebox, a context-sensitive, zero-input playlist generator that recommends and plays songs from the user’s personal MP3 collection. Initial experiments suggest that our system is more accurate than both a random generator and a system that does not take context into account.