CC-interop: COPAC/Clumps Continuing Technical Cooperation. Final Project Report

As far as is known, CC-interop was the first project of its kind anywhere in the world and still is. Its basic aim was to test the feasibility of cross-searching between physical and virtual union catalogues, using COPAC and the three functioning "clumps" or virtual union catalogues (CAIRNS, InforM25, and RIDING), all funded or part-funded by JISC in recent years. The key issues investigated were technical interoperability of catalogues, use of collection level descriptions to search union catalogues dynamically, quality of standards in cataloguing and indexing practices, and usability of union catalogues for real users. The conclusions of the project were expected to, and indeed do, contribute to the development of the JISC Information Environment and to the ongoing debate as to the feasibility and desirability of creating a national UK catalogue. They also inhabit the territory of collection level descriptions (CLDs) and the wider services of JISC's Information Environment Services Registry (IESR). The results of this project will also have applicability for the common information environment, particularly through the landscaping work done via SCONE/CAIRNS. This work is relevant not just to HE and not just to digital materials, but encompasses other sectors and domains and caters for print resources as well. Key findings are thematically grouped as follows: System performance when inter-linking COPAC and the Z39.50 clumps. The various individual Z39.50 configurations permit technical interoperability relatively easily but only limited semantic interoperability is possible. Disparate cataloguing and indexing practices are an impairment to semantic interoperability, not just for catalogues but also for CLDs and descriptions of services (like those constituting JISC's IESR). Creating dynamic landscaping through CLDs: routines can be written to allow collection description databases to be output in formats that other UK users of CLDs, including developers of the JISC information environment. Searching a distributed (virtual) catalogue or clump via Z39.50: use of Z39.50 to Z39.50 middleware permits a distributed catalogue to be searched via Z39.50 from such disparate user services as another virtual union catalogue or clump, a physical union catalogue like COPAC, an individual Z client and other IE services. The breakthrough in this Z39.50 to Z39.50 conundrum came with the discovery that the JISC-funded JAFER software (a result of the 5/99 programme) meets many of the requirements and can be used by the current clumps services. It is technically possible for the user to select all or a sub-set of available end destination Z39.50 servers (we call this "landscaping") within this middleware. Comparing results processing between COPAC and clumps. Most distributed services (clumps) do not bring back complete results sets from associated Z servers (in order to save time for users). COPAC on-the-fly routines could feasibly be applied to the clumps services. An automated search set up to repeat its query of 17 catalogues in a clump (InforM25) hourly over nearly 3 months returned surprisingly good results; for example, over 90% of responses were received in less than one second, and no servers showed slower response times in periods of traditionally heavy OPAC use (mid-morning to early evening). User behaviour when cross-searching catalogues: the importance to users of a number of on-screen features, including the ability to refine a search and clear indication that a search is processing. The importance to users of information about the availability of an item as well as the holdings data. The impact of search tools such as Google and Amazon on user behaviour and the expectations of more information than is normally available from a library catalogue. The distrust of some librarians interviewed of the data sources in virtual union catalogues, thinking that there was not true interoperability