DC/SLA

"Catalogers' Forecast: Metadata, Taxonomies, and Trends"

DC/SLA Event: Tuesday, February 4, 2003
Contact information for Anne Marie Houppert:
National Geographic Society
Voice: 202-828-6613 | e-mail: ahoupper@ngs.org

Speakers:
Rebecca Guenther, Library of Congress
Denise Bedford, World Bank
Suzanne Pilsk, Smithsonian Institution

Links to Handout and Presentations:

Bibliography/Recommended Sites [PDF] - Compiled by Anne Marie Houppert
Metadata Object Description Language Schema (MODS) [PPT] by Rebecca Guenther
Taxonomies for Information & Knowledge Management Architectures [PPT] by Denise Bedford
Getting that Gray Matter Working: Thinking Outside the Box of Traditional Cataloging [PPT] and XML Example [PDF] by Suzanne Pilsk

Summary write-up by Elaine Donnelly, National Geographic Society:

The law firm of Morgan Lewis Bockius and librarian Barbara Folensbee-Moore hosted a lively and well attended program (about 90 attendees) about cataloging trends. A reception preceded the panel discussion. The program organizer, Anne Marie Houppert of National Geographic Society, introduced the speakers.

Rebecca Guenther, of the Library of Congress' Network Development and MARC Standards Office, spoke about metadata in general and the Metadata Object Description Schema (MODS) project in particular. Metadata may come in different forms: administrative, structural, rights, preservation, and descriptive metadata. MARC, Dublin Core, and MODS are all types of descriptive metadata, but there are many others used.

MARC was developed as a communications format for bibliographic data 30 years ago and accommodates a wide range of bibliographic data with its many data elements. Dublin Core Simple, recently approved as ISO standard 8601, was designed to be a simpler set of data elements that could be used by non-libraries and libraries. Dublin Core offers a different approach to MARC with regard to digital objects, but is not complex enough for many library purposes.

MODS is an XML-based metadata schema that is designed with library applications in mind, but is flexible enough for non-library uses. MODS' bibliographic element set is based on a subset of MARC (19 main elements vs DC's 15 or MARC's hundreds). It is more user friendly than MARC, and its hierarchical arrangement allows for the description of complex digital objects. MODS records can be embedded into digital objects or maintained in a separate database.

The MARC Standards Office has developed MARC to MODS and MODS to MARC translation tools and posted them on their website (http://www.loc.gov/standards/mods/). MODS is meant to complement rather than replace MARC, acknowledging libraries' need for a rich descriptive metadata system that is less complicated than MARC but more sophisticated than Dublin Core. LC is using MODS for a number of projects including archiving of websites, an AV prototype project for digital preservation of audio and video objects, and for the Open Archives Initiative.

Denise Bedford, Senior Information Officer for the World Bank, talked about taxonomies. It is her opinion that even the professional literature presents a confusing and often simplistic view of taxonomies. A common misperception is that taxonomies are always hierarchical. There are several different types of taxonomies—some are hierarchical and some are not.

The four types of taxonomies are flat, hierarchical, faceted and network. To put it briefly, a taxonomy can be simple like a list, hierarchical like a good outline, link disparate elements together like a cataloging record, or link related terms at different levels in a hierarchy like a thesaurus. We use taxonomies every day in our work, but often call them rules or record formats. Don’t be afraid of them!

Suzanne Pilsk is a cataloger with the Original Cataloging department for the Smithsonian Institution Libraries, an organization that serves an extremely diverse group of researchers, staff and visiting scholars, administrators, and the public. Their online catalog covers materials in 20 branches from New York to

Panama

, and uses a MARC-based system. With the addition of digital materials to the catalog, MARC’s usefulness has been stretched to its limit. She related some of the challenges she faced when special projects arose, such as digitizing books with embedded Dublin Core metadata, cataloging online exhibitions, or describing a collection of digitized trade literature (early direct-mail catalogs, sewing machine manuals, and the like).

Her next project, the BCA Taxonomic Literature project, involves digitizing the Biological Centrali-Americana Project books and making the data not only retrievable, but able to be cited. Because this is a scholarly publication that is often cited in biology and biodiversity research, the researcher must not only be able drill down to the exact fact s/he seeks, but also be able to drill back up to know in which volume, fasicle, part, and page the information was found. She’s still working on her strategy for this one and is interested to trying MODS for this purpose. Her conclusion: we have to reach beyond MARC to cope with new generations of stuff coming into our collections, whether they be physical or digital objects.

Special Libraries Association \| DC/SLA Chapter PO Box 287 \| Benjamin Franklin Station Washington, D.C. 20044	A thank you to the Special Libraries Association (SLA) for hosting our DC Chapter's Web Site and Discussion List.
webmaster \| \|