sponsored by the Library of Congress Cataloging Directorate
What Can the Library Community Offer in Support of Semantic Interoperability?
Mary Charles Lasater
Nashville, TN 37240
The Conference topical discussion groups are for the purpose of identifying recommendations made by the speakers and commentators in their presentations and for developing recommended actions and an overall action plan for discussion and approval by the Conference in its concluding plenary session. Each topical discussion group consists of a facilitator who, with a designated number of participants, is assigned a specific topic related to the presentations that will serve as the focal point for identifying recommendations and deriving recommended actions and an overall action plan. Each group will also have an LC staff member to take notes and capture highlights throughout the discussion.Topic:
Several of the papers presented at this Conference refer to interoperability as a key issue in the identification and retrieval of networked resources. In his paper on Z39.50, William Moen underscores the importance of interoperability in achieving successful searches, and articulates the particularly difficult challenge presented by semantic interoperability. . According to the report of the National Academy of Sciences (NAS) LC21: A Digital Strategy for the Library of Congress (http://www.nap.edu/books/0309071445/html/) it is in the area of semantic interoperability where the knowledge organization and classification expertise of LC and the library community in general have the most to offer. By semantic interoperability, the report is referring to the extent to which different metadata schemes express the same semantics in their categorization (i.e., their labels for various fields). For example, when do elements have the same meaning? Is Principal Investigator , one of elements in the TEI (Text Encoded Initiative) metadata scheme, the same as Creator in the Dublin Core scheme and Author in AACR2 or, rather, is it equivalent to them? And what is meant by equivalence? These predicaments illustrate that semantic agreement among elements is required if users are to successfully and seamlessly discover and access networked resources across multiple domains. In the mid 1990s, mappings or cross-walks were created to determine this agreement. These were a useful mechanism at the time as there were essentially MARC and only a few prominent metadata schemes in place. In the past half decade, this situation has changed dramatically with the emergence of the EAD (Encoded Archival Description), the VRA (Visual Resources Association) Visual Document Description Categories, and the DGM (Digital Geospatial Metadata), among an ever growing multitude of specialized schemes. How many more such schemes will be developed we can only speculate, but the phenomenon has made it difficult to continue to apply this mechanism effectively. The problems are many and varied, but may be usefully identified as: structural interoperability, syntactic interoperability, and semantic interoperability. All three areas have involved the participation and cooperation of the library and metadata communities. It is clear that considerable work remains to be done on the problematic issues surrounding metadata interoperability.
Another kind of semantic interoperability (i.e., semantics of the contents of fields) is expressed by Lois Chan. In her paper on exploiting library subject and classification tools to retrieve networked resources she cites the prime importance of interoperability in enabling users to search among resources from a multitude of sources generated and organized according to different standards and approaches. Barbara Tillett in her paper on authority control on the Web notes a proposed experiment within IFLA (International Federation of Library Associations) that will require the exploration of ways to provide interoperability across multiple authority files. Thomas Mann argues that LCSH provides the heavy lifting capabilities needed by research libraries. Perhaps controlled vocabulary can be used to bridge subject aspects of metadata schemes. Researchers at OCLC, particularly Diane Vizine-Goetz, are undertaking extensive research with DDC is attempting to provide a possible classification bridge. The NAS report also mentions that LC, and by implication the library community, has much it can contribute by virtue of its extensive cataloging experience and its unique understanding of entity-relationships that are at the root of interoperability mechanisms. Can LCSH, LCC and DDC be used to enhance resource organization and discovery on the Web? How can known authority control mechanisms be used or enhanced to provide "interoperability"?Assignment:
In this assignment we are asking you to develop a list of 4-6 recommended approaches that the library community from its unique perspective could offer towards advancing semantic agreement among multiple domain-specific metadata schemes. Such approaches might exploit library tools, resources, and practices. They might also iterate the use of mechanisms to further knowledge representation or to extend entity relationship modeling. A list of approaches would be of obvious value to the metadata community, particularly such groups as the Dublin Core and the INDECS communities who are seeking common ground in the expression of metadata for networked resources. The list could also serve in the development of common standards for metadata interoperability. From the library perspective, such a list could serve to affirm the profession's leading role in the development and promotion of resource description and discovery standards.Procedures:
Your topical discussion group is organized into two parts to cover the two Conference days in which you meet.
Once you have finalized the prioritized list of criteria, the LC recorder will input it to a computer and a Powerpoint presentation will be created for your facilitator to present to conferees.Presentation and Action Plan:
Your facilitator will present the prioritized list of approaches for discussion and approval in the closing session of the Conference. Conferees will use this list along with the prioritized recommendations presented by the facilitators of the other topical discussion groups to develop an overall action plan that LC can carry forward from the Conference.