MODELS: MOving to Distributed Environments for Library Services
MODELS is a UKOLN project which receives support from the Electronic Libraries Programme. ELib is providing funding for four workshops and technical consultancy across the project. The salaries of UKOLN staff are covered by the core grant, which is provided jointly by ISSC and BLRIC. The project officially commenced on 1 August 1995, and will end on 31 January 1997. There will be a total of five workshops; the final one will be funded primarily by BLRIC. The project has been described as ‘groundbreaking’ and ‘making history’. It has already attained significant results in suggesting ways forward in managing distributed services and there is evidence that it is influencing developments in the field. A major achievement so far has been the organisation of three extremely successful workshops; very positive feedback has been received from participants. Two more are scheduled before completion of the project, although there has also been a proposal for a further informational event to present the final results to a wider group.
The first workshop was held in December 1995 in Bath. Its focus was the discovery, location and request of journal articles. It was partly educational and provided a very useful basis for the achievements of the other workshops. A number of important issues arose, stemming mainly from the current disjointedness of services and the differing levels of metadata content. In keeping with its role as a supporting study, the main results from MODELS 1 were a number of recommendations for further work to be carried out. These include a series of studies to critically evaluate specific standards for serials and recommend use within eLib; also a meeting to promote the use of serial part identifiers, such as SICI (Serial Item and Contribution Identifier). The scope and timescale of the studies is currently being planned by eLib and UKOLN. An email discussion list was set up at UKOLN to keep people up to date with further developments and provide an ongoing discussion forum. Given the shared interests which emerged between workshops one and three, it was decided to add workshop three participants to the same discussion list.
The second workshop took place in Warwick as it was a larger event and required facilities such as networking. It was also spread over three days unlike the others, which covered two days. As the subject was resource discovery and metadata, it was organised in conjunction with OCLC and functioned as a follow-up to the 1995 metadata workshop in the US, which produced the Dublin Core. Therefore another difference was that in addition to the UK participants, there were a large number of experts from North America, as well as several from Australia and Europe. MODELS funding covered only the expenses of UK participants; others were charged a fee. This workshop had a clear technical development focus. The ‘Warwick framework’ was the most significant result: a ‘framework for the deployment of resource description’. It recognises that a variety of metadata types will be required by different communities with different requirements, and seeks to provide an architecture to enable interchange between them. Reaching consensus between a number of leading implementors was also a notable achievement. The workshop report was published in Dlib magazine . A number of other useful papers resulting from the workshop also appeared in the same issue of Dlib. In addition, much useful work has been carried forward via a very active email discussion list.
Workshop three was similar in format to the first, and was also held in Bath to help reinforce the location of the project at UKOLN. It was useful in crystallising issues arising in the first two. The outcomes of this workshop are particularly significant and potentially groundbreaking. A focus of discussion was the concept of ‘clumping’. Clumps were taken to be aggregations of catalogues; these could be physical or (more likely in future), virtual. Other issues discussed included collection descriptions, and referral services to the descriptions. A three-tier architecture model based on requirements and standards discussed was presented. The middle tier (between client and server layers) was a ‘broker’, which would offer Z39.50 functionality and a range of other services, depending on its ‘richness’. Perhaps the most tangible result of the workshop was the proposal for a ‘National Agency’ which would perform a range of functions relating to a national discovery system. This would include maintaining collection and technical profiles to which participating catalogues would have to conform. It is significant that the British Library has already committed funding to explore the concept further. UKOLN is drawing up terms of reference for a National Agency scoping study. A parallel activity will be the development of demonstrators.
There is a basic overview of the project available on UKOLN’s Web server. It is intended to develop this more fully. For example, we will be mounting some of the interim architectural models for distributed services. We have received various queries resulting from information provided on the Web pages. Several overseas visitors (Australia, Sweden) have come to UKOLN specifically to learn more about the project. A presentation on issues raised by MODELS as well as interim results, was given to library staff at CERN and the University of Geneva in May 1996. Information about workshop reporting follows in the next section. (v) Finance Details of financial expenditure to date are attached at Appendix A.
Confidential
The third project line was changed from the originally planned subject of Libraries and the Web, to National resource discovery. The reason was because it was difficult to find a sensible focus, as the Web is seen to be integral to any architectural solution. The alternative line was a suggestion from eLib. The running order of workshops four and five has been exchanged. The final workshop will now be resource sharing. It was realised that this topic is a more suitable one for drawing various issues together. A further change is in the method of reporting. It was originally planned to produce a full report after each workshop, with a consolidated version at the end. The first workshop report turned out to be very long and covered many generic issues relating to the project as a whole. In addition, the change in subject of the third workshop meant that it shared some themes with the first, and it appeared sensible to combine the two. This has therefore become the basis for the final project report. However, we will be mounting work in progress on the Web server. Also, a full individual report from workshop two was published in D-lib magazine, and we have produced a summary report for workshop three; we will do the same for workshop one. It was difficult to define a reporting structure before the project started; this solution appears to be the most suitable. As shown in Appendix 1, we have some surplus of funding resulting from workshop two, as it was not possible to identify enough UK participants with suitable expertise. However, we have agreed alternative ways of using this. MODELS has benefited from being able to make some minor modifications. It is important to have some room for flexibility within a project. Without departing completely from the original aims, and with the agreement of the programme office, it should be possible to make sensible changes during its course. This could either be to take account of developments within the field, or to acknowledge interim results.
This section will be supplied separately, as discussed with eLib coordination. It is planned to include evaluation forms in participants’ briefing packs at the next two workshops. The positive feedback already received has been verbal and by email. Workshop two’s followup comments included: ‘I sense the workshop may have made history’(Tom Baker, German National Research Centre for Information Technology). Background briefing materials were produced for each workshop. Several participants have commented on the usefulness of these; extra copies of the packs prepared for the first workshop were requested for distribution to colleagues. Comments on this material will be requested in the workshop evaluation form.
There are two more workshops to be held, which will build on the work of the others. At the end of this time, we will be able to suggest an architectural model to manage the range of distributed library services covered in the project. It has been proposed that a one-day final roundup workshop would be useful to present the results and report to a wider cross section of the UK library community. The project has been making a number of suggestions for further work to be carried out in specific areas. Therefore many of the issues raised within MODELS will be continued in various ways. It would be desirable however to initiate some development work based on the ideas and models emerging; a certain amount of development is likely to take place resulting from the proposal for a National Agency in workshop three, but some wider work to cover a range of services would also be welcome. It was proposed at the national resource discovery workshop that an educational meeting for the wider library community should be held at the end of the project, to share the issues, outcomes and subsequent developments more extensively.
(i) Workshops The estimated cost of each workshop was £7,450, with an extra 15% allowed for workshop two (£8,567), given its extra length and requirements for networking etc. However, for the reason stated in Section 2, the actual expenditure for workshop two was less than anticipated.
Workshop 1: £7,553
Workshop 2: £5,406
The final figure for workshop 3 (held in July 1996) is not yet available, but it is anticipated to be close to the estimated amount. Therefore we currently have a surplus of approximately £3,000. It was agreed by JISC that some of this could be used to supplement the British Library’s partial funding of workshop 4. It was also agreed that a further portion could fund the participation of Sebastian Hammer (Index Data, Denmark) at the related Z39.50 event UKOLN is organising at Libtech.
(ii) Consultancy
Expenditure on technical consultancy from Fretwell-Downing is as anticipated.