4. Summary and Conclusion
4. Summary and Conclusion | |
Historical fashion objects exhibit rich human intelligence and knowledge, and illustrate humankind's history through fashion styles, colors, patterns, materials, and underlying techniques and design ideas. In preparing a "digital collection" (a database or a website) to document these three dimensional objects, a traditional method-- cataloging - was employed to record the attributes of a fashion item in a structured way. The unique characteristics of the fashion objects were examined and the way the data is accessed, presented, and researched by the known and potential users were analyzed. A digitized fashion collection on the Web creates a completely different dynamic for a material culture collection, overcomes conservation problems, and makes these artifacts available to large, and sometimes new, audiences. But, digitizing is not merely an appropriate and desirable alternative for museum exhibitions. The ability to search and browse digitized collections by various access points will greatly enhance the usefulness of the collections. To overcome the weaknesses of lacking search functions and evaluation support in current digitized fashion websites, users' needs and cataloging formats were analyzed. A set of descriptive elements were identified and grouped into three categories: Registration Information, Descriptive Information, and Subject/Topic Information. Because only a handful of digital collections of such objects exist and none of them have followed a standard metadata format, the elements identified in this project were compared using three metadata formats: USMARC, VRA Core, and Dublin Core.
Museum objects have characteristics that differ from information packages in most other environment, and these affect the description in the surrogate record (Taylor, 1999, p.97). Existing metadata formats, such as those examined by the study, were primarily designed for document and document-like objects and for limited types of objects. They have not progressed enough to include much of the key information related to a historical fashion object. The analysis in this study reveals that USMARC's strength is that it provides an exhaustive format for description that allows a cataloger to present various kinds of notes and differentiate various subject heading courses, but its use requires a great deal of time and professional skill. Important fields for authorship, title, and publication information are difficult to apply because the museum objects are often imperfectly known at the time of accession and registering. On the other hand, Dublin Core provides an easy application to create minimum level records. However, to apply the Dublin Core to the fashion collection without supplementing more descriptive elements would be inappropriate. VRA Core's treatment of a variety of specific elements matches fashion characteristics better than the other two formats. Consequently, VRA Core, supplemented with a few additional elements and modified data definitions, was applied and was found to be the most suitable to the digitized KSU historical fashion collection. Initial experiment also showed that the extended VRA Core might also be suitable for other three-dimensional objects, pending further testing.
In order to facilitate interoperability with other information resources which might have used or will use different metadata formats, tables to map the VRA Core format to USMARC and Dublin Core formats were developed. The process was costly and involved many complicated cases, partially because of MARC's high-level specificity and Dublin Core' high-level generality. VRA Core's principle that allows both original work and reproductions of the work be treated in one record contributed to the difficulty of mapping to metadata formats that are primarily designed for document and document-like objects. It is anticipated that mapping to additional metadata formats, such as the REACH Element Set and CDWA, may encounter complicated mapping process as well.
Fortunately, as the project progresses, there has been an increase in metadata studies. The most promising development is the syntactic foundation for Web-based metadata, the Resource Description Framework (RDF), which is now being developed under the auspices of the World Wide Web Consortium (Miller, 1998). RDF's intention is to encourage the exchange, use, and extension of metadata vocabularies among disparate information communities. The next stage of the author's research will follow RDF development and focus on the interoperability and compatibility of the modified VRA format with other information resources. While preparing the draft of this paper, it is also gratifying to note the release of the Encoded Archival Description (EAD) Document Type Definition (DTD) version 1.0. It is designed to function as both a Standard Generalized Markup Language (SGML) DTD and an Extensible Markup Language (XML) DTD. Although the current study does not test EAD for its suitability for creating catalog records for three-dimensional artifacts, its data model and structure provides valuable insights into mapping VRA into an RDF/XML-based scheme. It is hoped that the experience presented here will be valuable to other researchers involved in similar projects of metadata development and integration efforts.