Scherzo: a FRBR-based music discovery system
Mark Notess, Jon Dunn, Juliet L. Hardesty · 2011
The Scherzo music discovery system is one deliverable from the Variations/FRBR (V/FRBR) project at Indiana University (Riley, 2010). The objective of the V/FRBR project is to provide a real-world test of the Functional Requirements for Bibliographic Records (FRBR) model (IFLA, 1998) in the domain of music. In addition to creating a schema and FRBRization algorithm to populate a repository with data drawn from MARC bibliographic records, one experiment in the utility of the FRBR model has been to create a discovery system based on the FRBRized data, to explore the value of exposing FRBR structuring in the discovery interface. 2. Design and Implementation Scherzo is a web-based discovery layer atop a FRBR-based metadata repository (see FIG. 1). Metadata were generated by running the V/FRBR FRBRization algorithm against approximately 190,000 MARC records exported from the Indiana University Libraries ’ catalog. The selected records represent nearly 85,000 sound recordings and 105,000 music scores held by the IU Cook Music Library. The FRBRization process attempts to identify musical works from the bibliographic records, as well as identifying the people or groups responsible for the works, their expressions, or their manifestations. The FRBRization process also takes advantage of access to