The SALAMI project will be the first large-scale analytic application based on the NEMA (Networked Environment for Music Analysis) project. NEMA received its initial Phase I funding ($1.2M) from the Andrew W. Mellon Foundation in January 2008. Phase I is funded until 31 December 2010. NEMA is a multinational, multidisciplinary, cyberinfrastructure project for computational music analysis.  NEMA is a collaboration of six top laboratories in MIR and CM research: Goldsmith, University of London (Tim Crawford), McGill University, CA (Ichiro Fujinaga), Queen Mary, University of London, UK (Mark Sandler), University of Illinois at Urbana-Champaign, US (J. Stephen Downie), University of Southampton, UK (David De Roure), and University of Waikato, NZ (David Bainbridge).
The NEMA team is creating an open and extensible Webservice-based resource framework that facilitates the integration of music data and analytic/evaluative tools that can be used by the global MIR and CM research and education communities on a location- and time-independent basis. It is important to note here that NEMA is an infrastructure project, not an analytic project. Mellon is funding NEMA to develop and build out the underlying Webservices that will assist scholars, researchers, and students in their musicological studies. The SALAMI project proposed here is an analytic project that will, of course, build upon and exploit the NEMA infrastructure. It is as an analytic project that the SALAMI team is seeking funding from the DID Challenge. We further hope that SALAMI’s success will be instrumental in the NEMA project acquiring its Phase II funding from Mellon in 2011.
The NEMA project itself is based in part on previously developed technologies by the NEMA researchers including M2K (Music-to-Knowledge)  (Downie et al. 2005), Maestro  (De Roure et al. 2005), OMEN (On-demand Metadata Extraction Network) (McEnnis et al. 2006), and jMIR (McKay and Fujinaga 2007). These pre-existing technologies will be used as part of the SALAMI analytic work.
The NEMA framework is built upon a data flow execution system called Meandre (see Sec. F) , which is the product of another ongoing Mellon-funded project being conducted at UIUC called, SEASR (Software Environment for Advance Scholarly Research).  Since SALAMI will be using the SEASR/Meandre framework as its computational infrastructure, we will be in constant contact with the NCSA’s Automated Learning Group’s SEASR/Meandre development team for software support and advice.  We are eager to work with them to optimize SALAMI’s data management and I/O issues. We will also be working with the SEASR/Meandre team as we develop the Webservice data delivery and processing systems promised in Sections C.1, C.3 and C.5.
Each of SALAMI’s three laboratories is well equipped with multi-terabyte storage servers and several high-powered computational clusters upon which the team will develop the analytic software and run preliminary experiments (IMIRSEL: 96 cores, 30TB disk space, 376GB RAM; McGill: 36 cores, 14TB disk space, 76GB RAM; Southampton: ~1000 cores, 30TB disk space, ~900GB RAM).
In June 2009, the SALAMI team was awarded a grant of 250,000 normalized hours of supercomputing time at NCSA (National Center for Supercomputing Applications). The NCSA proposal was, in fact, based upon the SALAMI DID 15 March 2009 Letter of Intent. SALAMI’s extraordinary good fortune in acquiring this supercomputing opportunity will ensure that SALAMI has enough computing resources to perform: a) the testing and fine-tuning of our candidate structural analysis algorithms; b) our finalized structural analysis data creation runs; and, c), our exemplar experiments. 
See Appendix A: Fig. 4 for overall infrastructure of NEMA.
See http://www.seasr.org/meandre/. See Appendix A: Fig.5 for an example of Meandre Workbench.
Michael Welge, the NCSA Co-PI of SEASR, sits on the SALAMI Advisory Board.
Dr. Alan Craig, liaison for the NCSA computation time awards, sits on the SALAMI Advisory Board.