Finding the Hidden Gems: Recommending Untagged Music
Ben Horsburgh, Susan Craw and Stewart Massie
In this paper we present a novel hybrid representation for Music Information Retrieval. Our representation is built by incorporating audio content into the tag space in a tag-track matrix, and then learning hybrid concepts using latent semantic analysis. We apply this representation to the task of music recommendation, using similarity-based retrieval from a query music track. We are interested in measuring the recommendation quality, and the rate at which cold-start tracks are recommended. We develop a new approach to evaluating music recommender systems, which is based upon the relationship of users liking tracks. Our hybrid representation is able to outperform a tag-only representation, in terms of both recommendation quality and the rate that cold-start tracks are introduced to the system.