J. Stephen Downie (Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign), David De Roure (Oxford e-Research Centre, University of Oxford) and Kevin Page (Oxford e-Research Centre, University of Oxford).
Music Linked Data Workshop, 12 May 2011, JISC, London.
1. Towards Web-Scale Analysis of Musical Structure David De Roure J. Stephen Downie Kevin Page Ichiro Fujinaga Tim Crawford Ben Fields David Bretherton … salami.music.mcgill.ca
2. SALAMI Objectives SALAMI == Structural Analysis of Large amounts of Music Information Musical analysis has traditionally been conducted by individuals and on a small scale Computational approach, combined with the huge volume of data now available, will Deliver substantive corpus of musical analyses in common framework for music scholarsand students Establish a methodology and tooling so that community can sustain and enhance this resource www.diggingintodata.org
3. Motivation A resource of this size empowers musicologists to approach their work in a new and different way, starting with the data, and to ask research questions that have not been possible before The analysis is useful in classifying different genres of music and can be used to compare different styles of composition within a composer’s works or between composers It can also be used to understand historical influences over time and location
4. Digital Music Collections 23,000 hours ofrecorded music Music InformationRetrieval Community Community Software Student-sourced ground truth Supercomputer Linked Data Repositories
10. MIREX Overview Stephen Downie Music Information Retrieval Evaluation eXchange Began in 2005 Tasks defined by community debate Data sets collected and/or donated Participants submit code to IMIRSEL Code rarely works first try Huge labour consumption getting programs to work Meet at ISMIR to discuss results www.music-ir.org/mirex
13. Structural analysis processing time by different algorithms Evaluations of 3 algorithms and human against a ground truth FPC = Frame Pair Clustering