Unveiling Music Structure Via PLSA Similarity Fusion

Jerónimo Arenas-García, Anders Meng, Kaare Brandt Petersen, Tue Lehn Schiřler, Lars Kai Hansen, Jan Larsen

AbstractNowadays there is an increasing interest in developing methods for
building music recommendation systems. In order to get a satis-
factory performance from such a system, one needs to incorporate
as much information about songs similarity as possible; however,
how to do so is not obvious. In this paper, we build on the ideas of
the Probabilistic Latent Semantic Analysis (PLSA) that has been
successfully used in the document retrieval community. Under this
probabilistic framework, any song will be projected into a rela-
tively low dimensional space of "latent semantics", in such a way
that that all observed similarities can be satisfactorily explained us-
ing the latent semantics. Additionally, this approach significantly
simplifies the song retrieval phase, leading to a more practical sys-
tem implementation. The suitability of the PLSA model for rep-
resenting music structure is studied in a simplified scenario con-
sisting of 10.000 songs and two similarity measures among them.
The results suggest that the PLSA model is a useful framework to
combine different sources of information, and provides a reason-
able space for song representation.
KeywordsNMF, PLSA, Combining Music Similarity
TypeConference paper [With referee]
ConferenceIEEE International Workshop on IEEE International Workshop on Machine Learning for Signal Processing
EditorsKonstantinos Diamantaras, Tülay Adali, Ioannis Pitas, Jan Larsen, Theophilos Papadimitriou, Scott Douglas
Year2007    Month August    pp. 419-424
PublisherIEEE Press
ISBN / ISSNISBN: 1-4244-1566-7
Electronic version(s)[pdf]
BibTeX data [bibtex]
IMM Group(s)Intelligent Signal Processing

Back  ::  IMM Publications