Tailored tagger: A personalized, adaptive music tagging system

TitleTailored tagger: A personalized, adaptive music tagging system
Publication TypeUnpublished
Year of Publication2011
AuthorsMostafavi, A., Laurier C., Herrera P., & Ras Z. W.
Type of WorkISMIR late demo
AbstractWe present an adaptive music tagging system, the Tailored Tagger, which allows users to tag their music collections by mood and tracks the user’s tagging behavior. This allows the Tailored Tagger to alter the mood classification to clas-sify songs in the same way that the user would, resulting in a personalized mood recommendation system. To accom-plish this personalization, SVMs are retrained to correspond with the way the user tags. The result is a set of SVM clas-sifiers that are unique to each individual.
preprint/postprint documenthttp://mtg.upf.edu/system/files/publications/Mostafavi-ismir-2011.pdf