OpenURL Meeting
The JIBS User Group [1] holds regular workshops on issues relating to the use and development of electronic resources by the Higher Education community. The OpenURL was selected as a topic as JIBS perceived a growing interest in this issue, as shown by correspondence on email lists such as lis-e-journals, and the increasing uptake of OpenURL resolvers by the community. For example, the number of UK HE subscribers to SFX has risen from 5 in 2001 to 20 in 2003.
Three years ago, SFX was the only OpenURL resolver available, but the number of products has grown and there are now at least six. In this meeting, JIBS aimed to publicise the usefulness of the OpenURL technology and to provide information about selected products and how they differ.
UKOLN agreed to partner JIBS as they are involved in development of the JISC Information Environment technical architecture in which the OpenURL will play a role, and hence UKOLN also has an interest in explaining and promoting this technology to the community.
The meeting intended to achieve the following:
- to explain how OpenURL linking works, and define some of the terminology
- to enable a comparison to be made between different OpenURL products
- to benefit from the experience of institutions who have already implemented an OpenURL resolver
- to explore national initiatives concerning the OpenURL
Introduction to the OpenURL
Andy Powell of UKOLN set the scene by giving a very clear explanation of how the OpenURL works. He identified the problems associated with linking within the current Web environment, and demonstrated how OpenURL could provide a solution.
Librarians want to maximise the usage of our electronic resources - particularly journals - by a better joining up of resources, for example linking to the full text of articles from bibliographic databases. Proprietary linking services are unsatisfactory as they are usually dependent on business agreements between the database and journal suppliers; they give fixed URL links which do not take into account our preferred sources and sometimes lack currency. In addition, it is not economical in terms of Library staff time to attempt to support a plethora of different linking systems.
Libraries can set up an OpenURL resolver, or link server, which includes a local database (or ‘knowledge base’) of electronic journal subscriptions. This can replace proprietary linking systems: the library simply has to ask the suppliers of their bibliographic databases to put in a link which points to the OpenURL resolver. For this system to work, both the database (‘source’) and journals (‘targets’) must be OpenURL-compliant.
Commercial suppliers of OpenURL resolvers
Four suppliers were invited to present an overview of their OpenURL products: ExLibris (SFX); Endeavour (LinkFinder Plus); Fretwell Downing (OL2) and EBSCO (Link Source). Each was supplied with a brief, inviting them to explain how their product differs from that of their competitors, and in particular how the link server is set up and managed. They were also asked to give an indication of how much library staff time and expertise are required in setting up and running the product, and what type of training and support are available.
The majority of attendees found the product demonstrations useful, but a number commented on the repetitiveness of these presentations. Since the products effectively all perform the same task, this was difficult to avoid. This format however facilitated evaluation of the different products on offer for attendees who were at the early stages of reviewing the market place. Whilst there was the inevitable repetitiveness it was interesting to note the different emphasis and slants on the products that the different speakers made.
All products will perform the key task of providing a resolver, or link server, to join up bibliographic databases with full text or related information. There are differences in the quality of the knowledge base that is supplied with the resolver. A major problem still to be addressed is the accuracy of subscription data: neither suppliers nor publishers seem to know exactly what your subscription entitlements are! Data can be imported into the resolver from a third party such as Serials Solutions, or from a local Access database.
When considering which OpenURL resolver product to purchase, it is important to consider whether a ‘knowledge base’ is supplied with the resolver, and how well does it match your subscriptions. How many journal titles/publishers are included? Where does the data come from? How is it updated and how often?
Other questions to ask include:
- Is it a hosted or locally managed system?
- Is authentication required and how well does the resolver integrate with existing local and national authentication and authorisation mechanisms?
- How easy is it to import data? What import formats will it accept? Can this be done at publisher level or only title level?
- Which targets are pre-configured?
- Does it have the ability to generate an A-Z list of electronic journals for the library Web pages?
- Can it provide records in MARC format?
- Can you batch load data from the resolver into your Library catalogue?
- To what extent is local customisation and branding possible?
- How much library staff time and expertise are required, both in setting up and continuing to run the product?
- What type of training and support are available in the UK?
Some products have special features, for example a citation linker enabling users to input details of an article to try to find the text online, or a facility enabling the library to choose whether to link directly to full text from a database or to offer a range of menu options.
It was clear that you don’t need to use the same supplier for the OpenURL as for your library system: the OpenURL resolver runs independently of the library system. This undoubtedly has its benefits including freedom of choice of product, but it is a pity that the products cannot make use of the data already in our library catalogues, in order to populate the link server with data.
EBSCO was the only supplier to provide details of pricing, which is dependent upon numbers of FTEs (Full-Time Equivalents).
Implementers of OpenURL
There were three presentations in this section, from implementers of three different resolvers: SFX, 1Cate and LinkFinder Plus. The speakers were asked to provide information on why they had chosen the product; how much library staff time was involved in setting up and maintenance of the resolver; what training was provided; does the resolver assist in any way with existing library processes such as e-journals management or has it replaced existing processes; what sources and targets have been enabled and how well has the linking worked. It was clear from all of the speakers that the implementation of the resolvers was an opportunity to ‘expose’ their electronic resources and to contextualise them to the local environment for their users.
OpenURL resolvers had been implemented in order to:
- improve the use of e-journals
- provide seamless access for users
- provide a system for e-journal management
Owen Stephens from Royal Holloway noted that ‘SFX maintenance equals e-journal maintenance’: the task of keeping up to date subscription information still remains, but SFX has provided a system within which to manage this.
Successful linking depends upon:
- the quality of the subscription data in the link server
- the quality of metadata supplied in the OpenURL generated by the OpenURL source
A number of sources are not yet OpenURL-compliant. JIBS and other interested parties are lobbying JISC and CHEST on this issue, to try to make OpenURL compliance a condition of future licence agreements.
National initiatives
The Implementing the DNER Technical Architecture at MIMAS (ITAM) Project included the evaluation of SFX as a ‘National Default Resolver’. MIMAS set up a resolver which had no knowledge of the holdings of individual institutions but provided links to targets which are generally of interest to the community, such as Ingenta journals and Science Direct. This was tested by three participating Institutions: Plymouth, King’s College and Bath. Ross MacIntyre identified the ‘big plus’ of this approach as being ‘provision of new links to full text for users and independent of source service/publisher agreements’. MIMAS has hosted two OpenURL ‘talking shops’ which have provided a forum for UK HE institutions to discuss OpenURL issues [2].
EDINA are involved in two projects using OpenURL. Andrew Bevan introduced the Balsa linking system, which is the successor to EDINA links for linking from EDINA databases. EDINA have also proposed a project to JISC to set up an OpenURL router service. The intention is to route end-users seamlessly to their preferred OpenURL resolver (including a default national resolver if no institutional resolver is available) without having to contact each database provider separately to activate linking from their products. If successful, this would be extremely beneficial to the community as it would significantly reduce the institutional overhead of setting up OpenURL sources for their users.
Conclusion
Feedback from the meeting was very positive, for example: ‘Gave a clear feel of what can/cannot be done, what issues to consider, what work is involved…’. As with many technological advances it was clear that that technology was the easy part, the key factor for a successful local implementation in any environment is dependent on the accuracy and timeliness of the content of the so-called ‘knowledge base’. As we are all too well aware in the ever-increasing lean and mean budgetary reality, accountability for any expenditure is increasing. OpenURL resolvers may be viewed, to a certain extent, as enabling tools exposing resources and as such maximising investment in what are costly resources.The meeting was oversubscribed and JIBS/UKOLN have received several requests to repeat it.
The meeting did not cover all available OpenURL products. For example, WebBridge is in use at Swansea. A number of new products are being developed: the forthcoming resolver from Serials Solutions will be interesting since Serials Solutions are also able to supply subscription information. SIRSI have entered into a partnership with 1Cate and will be able to provide a resolver in the near future. TDNet have also developed an OpenURL product called TOUR; like Serial Solutions, this is an example of a product that links the technology to the content through its existing subscription-based knowledge base.
The OpenURL standard version 1.0 has recently been announced by NISO (National Information Standards Organization). This will extend the range of the OpenURL-based extended service that can be supported. For example, there is still no standard way of supporting ‘subject’-based extended services within the OpenURL resolver at present. The type of queries that can be used are dependent upon both the standard used and the source. We hope to keep a watching brief on developments in this area.
JIBS and UKOLN will consider providing another OpenURL event in future, which could include new product information and a discussion of developments resulting from the wider implementation of the OpenURL.
Presentations from this meeting are available on the Web [3].
References
- The JIBS User Group http://www.jibs.ac.uk
- MIMAS: ITAM: SFX Trial http://www.mimas.ac.uk/metadata/ITAM/sfx.html
- OpenURL Meeting, 17 September 2003: Programme
http://www.ukoln.ac.uk/events/openurl-2003/programme.html
Author Details
Frances Boyle
Electronic Resources Manager
Systems and Electronic Resources Service
Oxford University Library Services
Email: frances.boyle@sers.ox.ac.uk
Linda Humphreys
Faculty Librarian
Library and Learning Centre
University of Bath
Email: L.J.Humphreys@bath.ac.uk
Andy Powell
Assistant Director, Distributed Systems and Services
UKOLN
University of Bath
Email: A.Powell@ukoln.ac.uk
Web site: http://www.ukoln.ac.uk
Article Title: “The JIBS-UKOLN OpenURL Meeting”
Authors: Frances Boyle and Linda Humphreys
Publication Date: 30-October-2003
Publication: Ariadne Issue 37
Originating URL: http://www.ariadne.ac.uk/issue37/openurl-rpt/