National Library News November 1998 Vol. 30, no. 11

Virtual Canadian Union Catalogue (vCuc) Pilot Project Completed

by Carrol D Lunau,
vCuc Pilot Project Coordinator,
National and International Programs

The vCuc pilot project (see "The Virtual Canadian Union Catalogue (vCuc) Pilot Project", National Library News, vol. 29, no. 2, February 1997, pp. 13-14) was completed on March 31, 1998. The project involved 21 large libraries and 20 smaller libraries representing all types of libraries in eight provinces which use Z39.50 ¹ to search 18 servers on 10 different hardware platforms. Two-thirds of the participants had little or no prior experience using Z39.50 systems. This was one of the first projects in the world to involve so many library staff at all levels in searching such a broad cross-section of servers.

It was inevitable that some project participants would experience difficulties and initially react negatively. Two of the project's goals were to identify technical issues in using Z39.50 and to seek solutions to the problems encountered. As a result of this project, issues related to the transfer of holdings information found through a Z39.50 search, and difficulties created by varying definitions of keyword searching among different vendor products have been brought to the attention of the Z39.50 Implementors' Group (ZIG) ². The ZIG is in the process of defining an OPAC/Holdings Schema (see http://www.loc.gov/z3950/agency/) and identifying common attributes for keyword searching (see www.lac-bac.gc.ca/iso/z3950/keyword1.htm).

Despite the acknowledged current weaknesses of various Z39.50 products, users have high expectations. Seventy percent of respondents to the final project evaluation questionnaire anticipate that use of these products will change and improve the way they do their work. Eighty-one percent indicated that, if an operational vCuc existed, they would use it frequently for copy cataloguing, interlibrary loan, collection development, retrospective conversion, or reference. Specific benefits noted by the participants include:

• access to more locations for interlibrary loan, since not all libraries report to a centralized union catalogue or, if they do, holdings are not always current;
• access to more sources for cataloguing copy;
• "free" MARC records from most target databases;
• fast, inexpensive searching of several catalogues simultaneously;
• ability to search many catalogues using the same commands; and
• elimination of the cost of producing regional or local union catalogues.

Problems that users encountered during the project included excessively large result sets and unexpected results caused by different interpretations of attributes by different systems. During the project, a draft profile, defining a minimum subset of attributes (search indexes) and attribute combinations, was prepared. If all vendor systems used a common profile, interoperability problems between systems would be simplified and user difficulties minimized. The draft profile has already been discussed internationally, and the National Library will consult experts from Europe, Australia and the United States to develop an international profile. It will incorporate three levels of functionality: basic searching; item requesting; and catalogue updating.

The Directory of Z39.50 Targets in Canada was created for the vCuc pilot project by Scott Nickerson at Novanet. It contains information about a number of Z39.50 targets, giving the parameters for connecting to the system and for identifying the attributes and attribute combinations available for searching. The software has been provided to the National Library of Australia, which has created an Australian directory, and to the United Kingdom Office of Library Networking (UKOLN), which is using it as the core of a directory of U.K. targets.

The vCuc project demonstrated the viability of geographically dispersed catalogues, accessible via Z39.50 for copy cataloguing, and as a secondary search strategy, for interlibrary loan, with a small number of carefully selected targets. It will be some time, however, before it is technically feasible to operate within an environment in which a user can search an unlimited number of targets without being concerned about which product is used for the target, how the target has interpreted the attributes, and how the attributes have been mapped to the database.

Libraries are implementing Z39.50 systems at an unprecedented rate. The National Library will continue working with the Canadian and international library communities and with vendors to encourage the adoption of a common profile and the technical improvements being developed by the ZIG. Canadian librarians are encouraged to suggest to their vendors that they be early implementors of the OPAC/Holdings schema and the profile.

The National Library will also work to improve the reliability and performance of the existing AMICUS Z39.50 target. Once this has been accomplished and a pricing structure determined, a user support service for Z39.50 access to AMICUS will be developed. The National Library is now preparing a project proposal to investigate the creation of a gateway to Canadian Z39.50 targets on the Library's Web site.

The vCuc project concluded that Canadian libraries will continue to acquire and implement Z39.50 systems; however, coordinated development will not emerge without National Library leadership. Towards that end, the Library will develop a conceptual and technical model of vCuc which will be discussed with the library community as part of national resource sharing strategy discussions later this year.

The library community is in a transitional state between using highly centralized and widely distributed systems for access to library catalogues. The vCuc project has identified a number of issues which must be resolved before completely distributed systems will be effective and efficient. The balance between centralized and distributed systems remains to be determined. The present hybrid arrangement is likely to continue for some time. It gives libraries the choice of meeting local needs by searching a centralized database and/or searching a virtual catalogue. Most libraries will use both options, but the order and priority of selecting one or the other depends on such factors as the cost and ease of searching, and on existing consortia and partnership agreements.

Librarians and vendors now understand Z39.50 much better, but they still have further to go. Only by working together can they develop systems that will allow them to improve the services they can offer their clients -- the citizens of Canada!

______
Notes:

¹ Z39.50 is a standard for searching remote databases and returning result sets. It is now accepted by the International Organization for Standardication.

² ZIG is an international group of organizations implementing the Canada-U.S. standard.