Hubs and Homogeneity: Improving Content-Based Music Modeling

We explore the origins of hubs in timbre-based songmodeling in the context of content-based music recommendation and propose several remedies. Specifically, we find that a process of model homogenization, in which certain components of a mixture model are systematically removed, improves performance as measured against several groundtruth similarity metrics. Extending the work of Aucouturier, we introduce several new methods of homogenization. On a subset of the uspop data set, model homogenization improves artist R-precision by a maximum of 3.5% and agreement to user collection co-occurrence data by 7.4%. We also explore differences in the effectiveness of the various homogenization methods for hub reduction. Further, we extend the modeling of frame-basedMFCC features by using a kernel density estimation approach to non-parametric modeling. We find that such an approach significantly reduces the number of hubs (by 2.6% of the dataset) while improving agreement to ground-truth by 5% and slightly improving artist R-precision as compared with the standard parametric model.