Paper + Code(ISMIR2012): A Systematic Comparison of Music Similarity Adaptation Approaches

Sebastian and myself presented the Paper (pdf) at ISMIR 2012 in the first poster session of Tuesday, October 9th 2012.

Authors: Daniel Wolff(1), Sebastian Stober(2), Andreas Nürnberger(2), Tillman Weyde(1)

(1)City UniversityLondon, (2)Otto-von-Guericke-Universität Magdeburg

In order to support individual user perspectives and different retrieval tasks, music similarity can no longer be considered as a static element of Music Information Retrieval systems. Various approaches have been proposed recently that allow dynamic adaptation of music similarity measures. This paper provides a systematic comparison of algorithms for metric learning and higher-level facet distance weighting on the MagnaTagATune dataset. A cross-validation variant taking into account clip availability is presented. Applied on user generated similarity data, its effect on adaptation performance is analyzed. Special attention is paid to the amount of training data necessary for making similarity predictions on unknown data, the number of model parameters and the amount of information available about the music itself.

Download the poster here.

The code used in this publication is split along the collaborators into two parts. MLR and SVMLIGHT were tested using a framework developed by Daniel Wolff at the MIRG group. You can download the code and associated data using subversion from the following repository:

user / password: anonymous / citymirg

This entry was posted in research, tools and tagged . Bookmark the permalink.

Comments are closed.