In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • Sheet Music Consortium
  • Laurie J. Sampsel
Sheet Music Consortium. The Consortium, 2003. http://digital.library.ucla.edu/sheetmusic. [Requires a Web browser and an Internet connection.]

For information regarding the scope of this column, consult the headnote in the September 2006 issue (p. 155 of this volume). All Web sites accessed 22 November 2006 unless otherwise specified.

The Sheet Music Consortium (SMC) is a collaborative project providing access to multiple digital sheet music collections through a single database and search engine. It is a free resource, hosted by the University of California, Los Angeles (UCLA). The SMC uses the Open Archives Initiative (OAI) Protocol for Metadata Harvesting, version 2.0 (see http://www.openarchives.org).

Rather than visiting individual sheet music digitization project Web sites at various libraries, the SMC allows users to search several (currently seven) collections at once. The Dublin Core (DC) data for the collections is harvested from the data providers and housed in a database at UCLA. Links in the DC records, however, point to the digitized images residing at the various libraries' servers. The SMC content is not unique, but the search engine is. In effect, it is similar to a Web meta-search engine such as Dogpile, which markets itself as "all the best search engines piled into one" (http://www.dogpile.com). In comparison, the SMC could be described as all the best sheet music projects piled into one.

Three pieces of software are required to make the SMC (or any OAI project) work. An OAI compliant data provider is required at the participants' end to allow the data to be harvested in the OAI/XML format. The second program, the OAI compliant data harvester used by the SMC to request the XML files, was created using the Java programming language at UCLA. The third part is the service provider that makes the data available to users. The SMC compiles the files in an Oracle database, which is accessible via a search engine also programmed by UCLA in Java. In addition to searching the database, the SMC allows the user to create and save virtual collections.

The project was spearheaded by three member institutions: UCLA, Indiana University, and Johns Hopkins University. Duke University joined the Consortium later. The Mellon Foundation assisted with funding. Work began on the SMC in 2001 and the Web site went live in 2003. Since then, three additional data providers have had their metadata harvested: the Library of Congress (LC), Maine Music Box, and the National Library of Australia. There are currently over 120,000 records in the database. Not all of those records, however, have links to digitized sheet music. This issue is explored below.

To date, the collections participating in the project mostly represent the largest digital collections of American sheet music, making it a valuable research tool for the scholar of this format. The National Library of Australia is an obvious exception, but it does also include some American imprints. The Library of Congress collection included is "Music for the Nation, 1870– 1885." This is not indicated anywhere at the SMC site, but is clear from a date-limited search. The addition of the other LC collections, especially the earlier "Music for the Nation, 1820–1860," with ca. 15,000 digitized titles, would be an obvious recommended addition. One wonders why LC has not made its other digitized sheet music collections available for harvesting.

The following chart shows the total number of records available from each library, along with estimates of the total number of titles digitized. Of the 120,000 total records, about 88,000 (or 73 percent) include images. [End Page 663]

* * *

How well does the SMC work? To a certain extent the answer depends on the consortium. The search engine, documentation, and update schedule is crucial to the SMC's success. However, the rest is up to the participants. The quality of the metadata and scanned images is most important. Additionally, the data providers need to advise the SMC if their images move or become unavailable. The metadata provided differs from one collection to the next. The SMC harvests data in the fifteen core elements of Dublin Core (see the...

pdf

Share