RDF-DC Interoperability Testbed Project

This activity is now closed. The information on this page is provided for historical purposes only.


Demonstrate cross-disciplinary metadata interoperability using RDF as an encoding format and RDFS as the means for declaring schemas. Dublin Core will be the base metadata element set. Other metadata element set namespaces will be linked to DC via sub-property inheritance, thereby affording cross-discipline interoperability without sacrificing the native namespace of any given element set.


The Internet Commons is more effective to the extent that information resources are discoverable, retrievable, and renderable with the fewest possible impediments. The linking idiom of the Web solves an important dimension of this problem, but cross-disciplinary metadata standards are required to increase semantic interoperability across languages, disciplines, and sectors. The Dublin Core addressses this need for general information resources, but namespace branding and the need for extensions to basic discovery metadata often make it desireable to manage metadata namespaces locally.

RDF Schemas allow the declaration of metadata elements in ways that allow retaining control and branding of their local DC-derived namespaces declaring elements as subProperties of a DC element where appropriate. Thus, amico:artist can be declared as a sub-property of dc:creator, aglis:title can be declared as a sub-property of dc:title, and so on. Of course, there will be elements in datasets that are not sub-properties of DC as well, but by using schemas to declare relationships that are manifest, metadata designers can satisfy the need for local management and element set branding while still ensuring that their data will be visible in the larger context of the Internet Commons


This project will bring together a substantial amount of data from different sectors in different countries, unified using RDF Schema declarations as described above. Participants will be recruited from the government sector, museums, the commercial sector, supragovernmental organizations, and education. The resulting database will comprise a testbed accessible to researchers and designers to demonstrate and experiment with an operational cross-disciplinary store. This testbed will also provide a tutorial by example that will help to promote this approach to enhancing cross disciplinary interoperability.


  • Phase 1
    • Complete by June 30, 2002
      • Identify participants
      • Define suitable RDF schemas and a handful of records to test record import and schemas
      • Code simple prototype
  • Phase 2
    • Complete by September 31, 2002
      • Build the full-size database and complete user interface functionality
  • Phase 3
    • Ongoing, based on interest and available resources
    • Accept additional participants to enlarge the scope and size of the testbed. A common style of schema declarations should permit esay infusion of additional datasets into the database.

Phase 1,2 Participants

  • An agricultural database from the Food and Agriculture Organization of the United Nations
  • An Australian government database (from Archives Australia)
  • The Amico art image database
  • Data from a substantial instructional database project (to be determined)
  • Data from an XMP PDF and an XMP Photoshop data store

Requirements for participants

The interoperability testbed team will need:

  • an RDF Schema for metadata, structured as described
  • permission to index a substantive collection of records, and make them available in the test system.
  • The scope, size, and terms of use of the contributed records is negotiable.

Testbed functionality

  • display search attributes for a given db or for the intersection or the union of relevant metadata element sets
  • enter a structured search and display result sets
  • display records rendered based upon schema-specified structure
  • illustrate underlying metadata element relationships
  • illustrate differential searching (results with and without merged schemas)


  • Web-accessible union database of records from the various participants
  • Presentations and demos at conferences
  • Journal articles describing the methodologies and technologies that support the testbed
  • An ongoing testbed for experimentation in the RDF and metadata communities
  • Joint press releases to enhance visibility of the constituent data sources and promote the respective technologies (RDF, DCMI, XMP)

Project team