Learning Latent Semantic Models for Music from Social Tags
Authors:
- 1 June 2008
- journal article
- research article
- Published by Taylor & Francis in Journal of New Music Research
- Vol. 37 (2) , 137-150
Abstract
In this paper we describe how to build a variety of information retrieval models for music collections based on social tags. We discuss the particular nature of social tags for music and apply latent semantic dimension reduction methods to co-occurrence counts of words in tags given to individual tracks. We evaluate the performance of various latent semantic models in relation to both previous work and a simple full-rank vector space model based on tags. We investigate the extent to which our low-dimensional semantic spaces respect traditional catalogue organization by artist and genre, and how well they generalize to unseen tracks, and we illustrate some of the concepts expressed by the learned dimensions.Keywords
This publication has 10 references indexed in Scilit:
- Introduction to Information Retrieval2008
- An Efficient Hybrid Music Recommender System Using an Incrementally Trainable Probabilistic Generative ModelIEEE Transactions on Audio, Speech, and Language Processing, 2008
- Modeling Semantic Aspects for Cross-Media Image IndexingIEEE Transactions on Pattern Analysis and Machine Intelligence, 2007
- Automatic genre classification of music content: a surveyIEEE Signal Processing Magazine, 2006
- Support vector machine active learning for music retrievalMultimedia Systems, 2006
- An Efficient Solution to Factor Drifting Problem in the pLSA Model2005
- Unsupervised Learning by Probabilistic Latent Semantic AnalysisMachine Learning, 2001
- 10.1162/jmlr.2003.3.4-5.993Applied Physics Letters, 2000
- Indexing by latent semantic analysisJournal of the American Society for Information Science, 1990
- A vector space model for automatic indexingCommunications of the ACM, 1975