Music Selection for Internet Radio
Alex Weiss
Department of Computer Science
University of Canterbury
Abstract
Music. Radio. The internet. Three very interesting fields which are linked in this project on internet radio stations. A comparative study of existing radio stations is performed. Weaknesses and problems in current internet radio stations, as exposed by the comparative study, are identified. We outline the general problems observed. In particular, focus is placed on automatic music selection. Our approach is based on the notions of music genre, popularity, repetition, catalogue coverage, style continuity and \music programme" gen- eration. A simple internet radio station has already been implemented. The station serves as a platform to test our algorithms.
