This 1-day conference was the third in a series of events organised by the Hybrid Libraries projects funded by JISC via the eLib Programme, and supported by the DNER. The prior two events had been held at the British Library, in November 2000, and Manchester Metropolitan University, in the previous week.
The event was late in starting due to heavy snow having delayed several of the delegates. Indeed, many of those who had intended being in Edinburgh had to call off altogether. The absent list unfortunately included Peter Brophy of CERLIM (who was therefore unable to give the planned paper on HyLife), and Stephen Pinfield of the DNER, whose place in the schedule was bravely filled by Michael Breaks, Librarian of Heriot Watt University, who also opened the event with a brief introduction.
Bruce Royan, Chief Executive of SCRAN (the Scottish Cultural Resources Access Network) then talked about the Hybrid Library idea, and gave a brief historical outline of the place of hybrid libraries in the development of the eLib programme, and where they stand in relation to other types of access to data on the web. In an entertaining address, he used a bizarre composite illustration – a rabbit with tiger stripes – to convey the idea that ‘hybrid libraries’ are the outcome of a form of genetic engineering. Extending this, we may find that the engineering turns full circle, so that we end up talking about ‘librarians’ once again – newly engineered to handle multimedia. And perhaps at the same time our traditional ‘gatekeeper’ libraries will be reengineered into gateways. Royan illustrated the challenges ahead for the hybrid library by displaying a succession of media types - a slide of a painting by Salvador Dali, a video clip of John Logie Baird talking about his invention, an interactive multiview image of the Botanical Gardens in Glasgow, and a rotatable image of a skull. Such things need to be made available via libraries. We need to be able to deal with a diversity of data-types, interfaces and skill levels.
Royan finished his presentation by suggesting an architecture of hybrid library services within the UK. At the top level is the DNER (now to be known as PORTFOLIO). Key to PORTFOLIO are the subject hubs and the Data Centres – portal sites offering deep linkage between resources. The CLUMPS services deliver mid-scale diversity, and provide a ‘stress test for Z39.50’ as a cross-search technology. Three of them act as virtual regional union catalogues, and one provides a virtual subject union catalogue. Then, at the local institutional level, we find hybrid library services.
Michael Breaks spoke next, describing the five different angles which he could see on the Hybrid Library: people, service, resources, organisation and learning.
Hybrid libraries are impacting upon library staff, who must be developed to cope with changing motivation, needs and expectations. Similarly, the advent of the hybrid library is affecting the types of service we deliver, most notably in interlibrary lending and document delivery, which are becoming much more important. At the same time, traditional circulation of books and journals will decline as more material is available over the network. The most striking example of the shift from traditional circulation is in the move from short loan or reserve collections to ‘electronic reserve.’
The core resources with which our libraries deal are also changing. Teaching materials are being delivered via managed learning environments. The increasing attention being paid in HE to the quality of teaching materials (assisted partly by the funding merger with FE) requires that libraries become involved in the development of these learning environments. At every stage, too, we must ask the economic question. Rarely, at present, is it possible to view the switch from print to electronic resources and services as providing any cost-saving.
As organisations, we are changing our business processes, for instance in order to accommodate distance learning. New partnerships are being developed, and even with publishers, with whom libraries frequently locked horns in the past, we are exploring new ways to the future.
Finally, changes in learning methods offer richer opportunities to the hybrid library, which needs to support learners as they ‘move out of the classroom’. A coherent interface for teaching materials is required, and the library should be central to it.
Ian Upton’s presentation was on Builder: the DIY guide. The Builder Project’s focus was upon the hybrid library within a single institution (the University of Birmingham). Builder’s aim had been to run a number of hybrid library projects, and then to demonstrate these to the community, in order to illustrate to other academic libraries what was possible. It was conceived as a process, not a product, and Upton made the point that hybrid libraries will change with changing technology.
A key concept for the project was ‘seamless access to resources’. The Builder team were developing by analogy with the campus swipe ID card, which can give access to CDs, training materials, the help desk, etc. In the hybrid library world, this implies authenticated access to services, which Builder has based on an ID card. The team has also sought to make the interface seamless (i.e. it only requires to be learned once), and maintains user profiles so that only appropriate selections of resources are presented to users.
Builder has produced a number of models to illustrate the components of the hybrid library. These are:
- Electronic short loan. The project digitised resources, cleared copyright and devised a way for students to access them securely.
- Exam papers database. Three years’ worth of University of Birmingham exam papers are now available on the web. Originally students obtained permission to access the database on the basis of their IP address, but now they are using a person-based authentication system. Ian Upton described this as a ‘frighteningly popular service’ (a recent log had shown 2,500 accesses in a single day).
- Forensic linguistics and Midland history. These are two journals published by the project. Articles are made available in PDF format and fully indexed. Subscribers can access the full text of articles.
- Hybrid Libraries search engine. This engine provides a search facility across most of the eLib Hybrid Library projects - Agora, Builder, Headline and Malibu. The search engine can index PDF, Word and Powerpoint files, as well as HTML pages.
- Thin Client demonstrator. This provides an interface to CD-ROM databases via a web browser from anywhere on the internet. The thin client technology means that the platform of the CD-ROM is irrelevant.
- TALIS toolbox demonstrator. This was developed in conjunction with Talis, which supplies Birmingham University Library’s library management system. The toolbox provides access to an Application Programmer Interface to TALIS, allowing the library to reuse TALIS data in various customised ways (for example, catalogue records can be extracted against profile information and presented to users alongside other personalised information).
- Induction Packages. These have been developed using Macromedia Flash. They teach novice users how to use certain library and information services by ‘popping up’ at appropriate points.
Upton finished his presentation by stating that the work of the Builder project was now informing the development of the University of Birmingham portal. Its iterative approach to development was somewhat at odds with the established culture, but its deliverables were sufficiently impressive to have gained widespread acceptance within the university.
Jessie Hay of the Malibu project then gave a presentation entitled Modelling the hybrid library: Aiming to give ’em what they want! Malibu has particularly focussed on the humanities area. Southampton, King’s College London and Oxford are the three major partners, working to produce ‘a generalisable framework and models’. It has created a small number of prototype hybrid libraries, including one for the Refugee Studies Centre at Oxford.
Malibu is particularly concerned with understanding user behaviour, and has held a number of end-user ‘walk through days’. These examined the different approaches to authentication in use in the partner institutions, and asked the question ‘at what point in the navigation should the challenge come?’ They also sought to find out how teachers were using resources, and what they were telling their students. Learning resources are considered to have a ‘cycle’ of use, and Malibu has been asking the question ‘at what points in that cycle should the library be involved?’ How has the advent of the hybrid library made library involvement possible now when it has not been in the past? This has required that the project engage in an exercise in deconstructing traditional library functions. Key to the whole new approach is staff development.
In contrast to Builder’s practical approach. Malibu has found the use of models very useful. Malibu has also produced a search engine prototype based upon identifying what resources people use, rather than starting with z39.50-compliant resources. The target environment included Google, OPACs, archives and a number of different data types. A ‘Global Information Gathering Agent’ communicates queries and requests with each target using the relevant protocol. Users liked being able to reprioritise their targets (e.g. to search a specific archive first, Google next, etc).
John Paschoud of the Headline project then gave a paper entitled Two million users? Two million libraries: The Headline PIE hybrid library model. The title is based on the fact that there are approximately two million users in UK HE.
Headline’s Personalised Information Environment (PIE) has become in many ways the focus of the project – though it did not start out that way. But Headline’s user studies found that everyone wanted something different from the hybrid library, and therefore personalisation became important. The project set out to achieve this by adopting the relational database concept of ‘views’ – with these being defined for users by means of profiling. Users frequently collaborate in their information seeking (i.e. they ‘hunt in packs’) and so a single view may be shared by a coherent group, such as students on a particular course. However, Paschoud explained that the PIE does not go far enough. Users want to share resources with each other and recommend resources to each other. This is where the ‘open environment’ being developed by Project ANGEL will be beneficial. ANGEL is one of the recently-funded DNER projects which builds upon Headline to some extent
Paschoud admitted to stealing ideas shamelessly from the commercial exploiters of the web. One of these is the personalised portal. ‘MyNetscape’ and similar offerings are rapidly become among the most popular web sites on the net, and achieve high degrees of customer loyalty. So Headline decided to follow suit. Paschoud explained that we know what users should be interested in from the courses they are on, and we know what rights of access they have. Personalisation can proceed from there.
Headline has also produced a number of other useful hybrid library products.
- ‘Sherlock’ is a database-backed product which takes users from the shelfmark to an item’s physical location.
- ‘Smart bookmarks’ (whose URLs don’t break) are managed by the PIE for its users.
- ‘Authbroker’ is middleware which manages the interface between the PIE and licensed resources. It is based upon an LDAP directory.
- RDM (Resource Discovery Metadata) modules are bolted on to the back of the PIE. These contain metadata for the PIE, managing discovery-to-access tools as well as resource metadata. The project proposed that RDM editing should be ‘mainstreamed’ with library cataloguing operations in order to ensure quality metadata and avoid duplication. The RDM modules also contain metadata describing how to search particular hosts, and serve the results.
Each of these developed packages are freely available open software, and – with their documentation - have been available from January 2001.
David Palmer of the AGORA project then described AGORA: From Product to Research – and gave a live demonstration of the product in action. The partners in the project have been the University of East Anglia, UKOLN, Fretwell-Downing and CERLIM – the Centre for Research into Library and Information Management at Manchester Metropolitan University. The project has used a group of Library Associates - Hull, Heriot-Watt, Bath Spa, East Anglia and Edinburgh (which joined later).
AGORA has commissioned Fretwell-Downing to develop a z39.50-based cross-search tool. Palmer admitted that it was ‘by no means a perfect product’, and that it has received some criticism. Nevertheless, it has steadily improved since the first prototype release, and Fretwell-Downing have adopted most of the suggestions made by the academic partners.
The concept of individualised ‘landscapes’ is employed. Resources are divided into those which are remote and those which are local. The product is designed to integrate with library’s interlibrary loan functions – a process Palmer described as ‘vertical integration’. Search results can produce a web form which is pre-populated and allows a request to be despatched immediately. Authentication has been a problem, however, and unfortunately it has not been possible to integrate Athens with the product’s authentication system in the time available.
The afternoon session was given over to workshops allowing delegates to try out the systems we had spent the morning hearing about.
This was an instructive day, despite the attempts of the weather to foil it. It is clear that a lot of good work has been produced by these hybrid library projects, and a number of tools have emerged which can be picked up and used by HE libraries across the country in rolling out their own hybrid library services. Predictably, a lot of work remains to be done – some of which is already underway in the latest round of projects funded by the DNER. The hybrid environment in which our libraries seek to redefine themselves is hybrid in several ways: a hybrid of print and electronic, of open-source and commercial, and of information and learning. As a community we should be glad of the investment made by JISC in helping us both understand this new environment, and accommodate our services to it.