A Case-Based Song Scheduler for Group Customised Radio

This paper presents a knowledge-intensive Case-Based Reasoning system to generate a sequence of songs customised for a community of listeners. To select each song in the sequence, first a subset of songs musically associated with the last song of the sequence is retrieved from a music pool; then the preferences of the audience expressed as cases are reused to customise the selection for the group of listeners; finally listeners can revise their satisfaction (or lack thereof) for the songs they have heard. We have integrated this CBR system with Poolcasting, a social group-based Web radio architecture in which listeners can actively contribute to the music played and influence the channels programming process. The paper introduces the Poolcasting architecture, presents the CBR technique that tailors in real-time the music of each channel for the current audience, and discusses how this approach may radically improve the group-satisfaction of the audience for a Web radio.