Over the last year, the consortium that makes up CIMI (Computer Interchange of Museum Information) [1] has been testing the feasibility of using Dublin Core [2] metadata records to make the discovery of museum objects easier, particularly for the non-expert. Natural history interests on the project are represented by Jim Beach from The University of Kansas Natural History Museum and Neil Thomson from The Natural History Museum in London.

Dublin Core is a set of 15 metadata elements originally intended to facilitate the discovery of networked resources by increasing the precision of responses to enquiries in comparison to the catch-all approach of the large search-engines. It has caught the interest of communities such as museums, libraries, government agencies and commercial organisations in over 20 countries, and is now also being applied to non-electronic or "offline" resources such as museum objects.

The Natural History Museum [3] intends to make use of this emerging international standard to create a cross-domain enquiry system for its own collections of specimens, literature, artwork and archives, by harvesting Dublin Core records from the various rich-description systems that are used to manage these collections. The perceived advantage of using standards is that the system should be able to mesh with enquiry systems that are based on the same standards and also span institutional and geographic boundaries.

The CIMI project aims include the production of an expanded version of the standard Dublin Core User Guide. This will offer guidance to museums and galleries on which elements should contain which pieces of information, so that the chances of interoperability are enhanced. A second aim is the creation of a pool of Dublin Core records in XML (eXtensible Markup Language) so that the problems associated with record creation and subsequent retrieval can be assessed.

The project thus far has looked only at Unqualified Dublin Core. The use of qualifiers to add significance to each of the 15 elements relies on the use of RDF (Resource Description Framework) [4], an XML application specifically designed for carrying metadata. RDF was not in a sufficiently settled state to allow meaningful decisions to be made, but it is expected that phase II of the project, which is just starting, will be able to tackle the use of qualifiers. The Dublin Core itself is in the process of development to allow its use by those concerned with intellectual property rights [5].

For natural objects, the three elements TITLE, CREATOR and DATE together form a citation, and the three should relate to each other. It seemed that a choice would have to be made between using the name of the determiner or the collector as the CREATOR for Unqualified Dublin Core records. However, CIMI has proposed the introduction of a new TYPE, which is the EVENT. It became clear quite early on in the project that the TYPE element held the key to the correct interpretation of a Dublin Core record, and for natural objects, it will allow records for both naming events and collecting events to be created - each with its associated CREATOR and DATE.

For phase II, CIMI is combining three of its testbed projects (Dublin Core, Z39.50 and Integrated Information Management) to concentrate on the use of RDF to carry Dublin Core metadata, and the retrieval of those records using the Z39.50 standard.

Further information can be found at:

[1] http://www.cimi.org/ [2] http://purl.org/DC/ [3] http://www.nhm.ac.uk/ [4] http://www.w3.org/RDF/ [5] http://www.dlib.org/dlib/january99/bearman/01bearman.html

 Neil Thomson
 Head of Systems & Central Services
 Natural History Museum Telephone: +44 (0)171-938 9005
 Cromwell Road Fax: +44 (0)171-938 9290
 London SW7 5BD Email: N.Thomson@nhm.ac.uk

NHM Web Page: http://www.nhm.ac.uk/