Paul Miller, Archaeology Data Service (collections@ads.ahds.ac.uk)
This publication represents the culmination of a long and complex process involving over one hundred subject and technical specialists from around the world in eight workshops, electronic discussions on more than a dozen mailing lists, a comprehensive and inclusive consultation programme and the unification of diverse viewpoints and positions. As such, it has been constantly enlightening, frequently stressful, often hard work and, ultimately, extremely rewarding for those closely involved. Contained within this volume is a clear model for implementation of the Dublin Core metadata element set (Chapter 3) within the Arts and Humanities Data Service and beyond and, in Chapter 4, the proposed technical solution to integrating at least five diverse and distributed electronic catalogues is presented.
Given the background to this work, both of these initiatives are intended to meet the current and pressing requirements of the AHDS, but both models were envisaged from the outset as both extensible and of wider relevance to the broader community. UKOLN's close involvement throughout ensured that an awareness of the broader picture was never far away.
Having survived to reach this stage in the proceedings, it is perhaps worth pausing to reflect on a number of lessons learned during the process before highlighting a few of the issues which remain as yet unresolved. In many cases, solutions for these problematic issues will require community-wide involvement and consensus.
The first - and perhaps most important - point reinforced during the consultation process was that the ultimate goal of effective cross-domain resource discovery was felt to be important by the community; a fact stressed repeatedly during the MODELS workshop reported in Chapter 2. Although it was considered that cross-domain searches across all of archaeology, history, textual studies, and the performing and visual arts together were perhaps unlikely, many respondents saw a value in being able to search coherently through the holdings of two or three subject areas at a time.
Such a finding is of wider relevance than to the AHDS alone - the Consortium for the Interchange of Museum Information (CIMI) and the Research Libraries Group (RLG), amongst others, are both exploring cross-domain resource discovery initiatives (CIMI 1997, RLG 1997) - and illustrates a widely felt need for greater congruence between the vast number of excellent but isolated resources currently available to users in the arts and humanities and beyond.
When this programme was first mooted, many doubted the possibility of unifying resource discovery structures across five subject areas and numerous data structures and formats. It was even doubted that we could agree on a common framework within which to work.
The remarkable consensus achieved, and presented in Chapter 3, belies the doubters, and is a testament to the hard work of those involved. Not only did they bring the benefit of their great expertise to the process, but they were also prepared on more than one occasion to surrender previously jealously guarded aspects of disciplinary territoriality for the sake of the cross-domain whole.
It would have been relatively straightforward - and certainly far easier than the consultation process actually implemented! - for a small number of data service staff to develop their own idea of an appropriate resource discovery metadata framework and then impose this upon data holdings.
Despite being harder to undertake, our consultation process offered a far better end result, in that the result is more clearly thought out, it includes the benefit of diverse viewpoints, and it offers the possibility of community 'ownership'; the ideas expressed in this volume are, after all, an expression of the combined wisdom of a great many experts rather than a small number of individuals.
The recommendations offered herein are consensual and ultimately extensible. They will be implemented by the AHDS, and are offered back to the wider community within which they were created in the hope that other organisations, too, will make the most of all this hard work. The benefits of such widespread adoption are potentially great as, whilst not only allowing other data services to benefit from the processes outcomes, there is also scope for aggregation of numerous data services nationally and internationally, all sharing a common view of the Dublin Core and of resource discovery cataloguing.
Even once implemented, the structure for resource discovery metadata offered in Chapter 3 will continue to evolve as internal and external experience grows in actually providing an effective resource discovery service. Certain of these changes will be minor, with SCHEME and TYPE definitions added, removed, or refined from time to time, but others are more major, and carry more wide-ranging implications. Dempsey et al. address several such issues of relevance to metadata-related endeavours in general in Chapter 5, whilst a number of direct relevance to this implementation of Dublin Core are raised below.
Chapter 3 broaches the subject of registries for definitions and usage guidelines related to Dublin Core's SCHEMEs and TYPEs. The AHDS forms one logical grouping, and will maintain a central registry to which new options may be added for the AHDS' implementation of Dublin Core.
Other groups implementing Dublin Core - except in its purest form - are also likely to discover a requirement for such a system, and it will be important for guidelines such as those in Chapter 3 to evolve in a manner capable of handling growing complexity.
Given the existence of an AHDS registry, and registries for other communities implementing similar systems, the next logical step is for a single super-registry capable of representing the overlap between diverse discipline-specific registries.
The requirement for such a model is some way in the future, and may well be superseded by parallel initiatives such as Z39.50's Explain facility, or even Centroids. Nevertheless, the possibility of such a requirement should be borne in mind, especially in the light of proposed developments such as the UK's National Agency for Resource Discovery (Kay and Brophy 1997) which will strive towards the creation of UK-wide implementation frameworks.
The Warwick Framework (Lagoze et al. 1996) offers a model by which relationships between different metadata structures and resources may be expressed and managed. As such, it is an important aspect of any long-term strategy for providing access to metadata and data, and development beyond the conceptual stage should be encouraged however possible.
The basic structure of resource discovery metadata offered here is unlikely to alter significantly, but realisation of the potential of the Warwick Framework promises to greatly improve both the provision of more detailed metadata and the relationship between resource description and resource.
If all of the work carried out to date is to be of maximum value to the community, it is going to be important to raise awareness of metadata's importance still further. Much work has already been done in this area, with metadata in general and the Dublin Core in particular to be found entering the vocabulary of many disciplines.
The next stage for widespread adoption is that of implementation, with a concomitant requirement for wide-ranging guidelines on implementation. These should include such details as illustrative examples of the benefits offered to specific data types or subject groupings by effective recording of metadata, guidance on the identification of suitable information for recording about resources of different form, etc. These guidelines will serve to ensure consistency of application across implementing groups, leading to a corresponding confidence that resource discovery will function as expected within and between implementations.
In conclusion, the initiatives reported here serve as an important part of the ongoing drive towards seamless electronic discovery, evaluation, retrieval and use of diverse, distributed, scholarly resources. The long-term vision of an electronic window on the world from every desktop is brought a step closer, and a short-term requirement to make sense of the morass of poorly described, disorganised data already out there is largely solved. It is now time for the AHDS and others to test these solutions by implementation and, with the problems thus resolved, to move on to the next set of challenges which inevitably await us.
Send comments or questions to info@ahds.ac.uk
Last modified: Monday, 17-Nov-97 16:52:01 GMT by D. Greenstein
URL: http://www.ahds.ac.uk/public/arlist.html
|
This page was originally part of the Arts and Humanities Data Service (AHDS) Website: http://ahds.ac.uk/public/metadata/disc_08.html
|
|