In this paper we present PAPA, a novel approach for au-tomatically generating playlists. The proposed frameworkutilizes the user’s physiological response to music, togetherwith traditional song meta-data to generate a playlist theuser will not only enjoy, but which will assist him or herin achieving various user-defined goals (“purpose”). In ad-dition to outlining the generic framework, we present an ex-emplary application named MPTrain that (1) creates a playlistin real-time to assist users in achieving specific exercise goals;and (2) incorporates the user’s physiological response to themusic to determine the next song to play

PDF file

Details