The OPAL Project: Developing an Online Digital Reference Service for Distance Learners

Citation

Bradbury, D. and Payne, G. (2001), "The OPAL Project: Developing an Online Digital Reference Service for Distance Learners", Library Hi Tech News, Vol. 18 No. 9. https://doi.org/10.1108/lhtn.2001.23918iab.001

The OPAL Project: Developing an Online Digital Reference Service for Distance Learners

The OPAL Project: Developing an Online Digital Reference Service for Distance Learners

David Bradbury and Georgina Payne

The last few years have seen an almost exponential growth in the World Wide Web (WWW), in terms of both its size and the variety of applications that can now be found "online." Such a technology offers obvious opportunities for distance learning, enabling students to have access to a far wider range of resources than was previously available using more traditional means of communication.

The Open University (OU) is a higher education institution based on distance learning with around 200,000 students, both undergraduate and postgraduate, enrolled in distance learning courses. As the majority of OU students do not live within travelling distance of the campus, until the advent of the Web, the OU Library played a minor role in student studies. There is now the opportunity to use the library as a Web-based service to offer a wide range of information resources to OU students. To provide such a service the OU Library has developed Open Libr@ry, an information resources portal for OU distance learners. There are also plans to extend Open Libr@ry to incorporate personalised student access through a "My Open Libr@ry"-style interface.

With the opening-up of the OU Library to students comes a significant increase in the number of enquiries that have to be handled by the Library's Learner Support Team. It is clear that a large number of staff would be needed to handle the potential number of enquiries that could be sent by up to 200,000 students. The issue is complicated by the fact that 50 percent of OU Library enquiries are made outside office hours, either after 5.30p.m. on weekdays or during the weekend. These enquiries range from routine requests for information relating to journals, passwords and services, which make up around 60 percent of queries in one set of sample OU Library enquiry data, to complex and specialised subject information enquiries. Additional analysis of OU Library enquiries has also shown that around 60 percent of enquiries could have been satisfied by existing information held on the OU Library or course Web pages. Some form of 24/7 online automated system (a fully automated virtual assistant available 24 hours a day, seven days a week, to answer repeat and routine enquiries from students) that could answer a significant proportion of routine enquiries, perhaps also linking students to relevant information held on library Web pages, would help ease the burden on library staff, whilst also providing an out-of-hours enquiry service for students. It is the remit of the OPAL Project (Online Personal Academic Librarian) to research and develop such a system. Eventually, OPAL will become part of the Open Libr@ry site, providing the site's out-of-hours and routine enquiry service. The team working on the OPAL Project includes two dedicated research staff based in the OU Library and three further library staff, with additional expertise drawn from researchers at the OU Knowledge Media Institute (KMI), partners in the OPAL Project. The Knowledge Media Institute specialises in the development of near-term future technologies for sharing, accessing, and understanding knowledge, such as the use of organisational memory and the use of intelligent agent systems.

It is not only the OU and its distance learning students who need a 24-hour enquiry service. The issues encountered by the OU are also faced by other libraries in the public, academic and research sectors which support lifelong learners, part-time and distance students.

Accordingly, in addition to the OU Knowledge Media Institute, we are working on the OPAL project in co-operation with two partner institutions, Birkbeck College in London and the Distance Learning Unit at the University of Leicester Library. Both partner libraries offer services and resources that differ from those at the OU Library, and also have a different student user base, enabling the OPAL project team to develop a model of the OPAL system that can be adapted to a variety of institutions. It is planned to test the prototype OPAL system with students and staff at both Birkbeck and Leicester.

To further ensure the adaptability of OPAL, the project team has drawn on the expertise of information professionals working in the public, research and academic sectors, and not involved in the OPAL Project as partner institutions. In February a focus group of 13 information professionals were invited to the OU Library to discuss the patterns underpinning reference enquiries and to describe the functionality and capacity of an ideal automated reference enquiry system. In particular, the focus group emphasised the need for understanding the whole enquiry context, usually ascertained during face-to-face enquiries via the reference interview process. The enquiry context includes all the verbal and non-verbal clues that give the librarian the information they need in order to select the appropriate answer, such as the manner, language and terminology in which an enquiry is expressed. This context also includes the user's previous library experience, their level of information and IT literacy, expectations concerning the level of service available, cultural and professional background, the specific course being studied, and the urgency of the information need. All impact upon the answer selected by a real librarian, and where possible the OPAL system will need to be able to identify these differing user contexts by means of user profiling in order to deliver a more specific answer to the user. In the case of universities, much of the information required for user profiling is held on student authentication systems. Public libraries, with their particularly wide user base and remit, pose greater problems for user profiling, and represent a further area of research for the OPAL Project.

The OPAL Project is currently in an initial 18-month research stage, during which a prototype system capable of answering reference enquiries posed in natural language is being researched and developed. The project team is creating software to deal with some of the more routine questions, such as requests for passwords. Initial analysis of a data set of over 500 questions from the OU Learner Support e-mail archive shows that 30 percent of these questions are related to passwords. Thus simply being able to deal with passwords would significantly reduce the enquiry desk workload. It is anticipated that such enquiries can be dealt with by a selection of filters consisting of a series of trigger words. When a filter is triggered the appropriate answer would be generated and sent back to the student. Thus the system must be able to extract significant levels of information from the original message sent by the user in natural language text.

With such a Web-based system a variety of interfaces is possible. Three interfaces are currently under consideration for the OPAL prototype:

1. 1.

   E-mail-based. A student sends a message to the server via a standard e-mail message. Replies to the question are sent back via e-mail, and there may also be additional messages asking the user to clarify their enquiry or giving them an estimate of how long it will take to answer their question.

2. 2.

   Web-based. This system would take information from the user via a form on a Web page. The response would then be delivered to the same Web page, in much the same way as Web form-based applications at present. For this approach to be used it will be necessary for the system to have a consistently fast response time.

3. 3.

   Chat-based. Perhaps the most radical approach under consideration would be for the user to chat with OPAL in real time via a chat room. The time constraints mentioned for the Web-based interface will apply even more tightly to this last approach.

It is then hoped to move to an implementation stage, during which the system can be scaled up and integrated with existing university systems to aid user profiling and to help deliver answers more readily geared to the user's context. Finally, the team hopes to develop an "artificial librarian", a more advanced addition to the filter-based system. This would involve building a community of intelligent mobile software agents capable of traversing the library's network of databases in order to answer highly specialised and technical questions ­ theoretically any question that a librarian can answer could be answered by this system. Such agents would have to be mobile, so as to make efficient use of CPU resources across the network.

Over the lifecycle of the OPAL project a variety of technologies will be used, many of which, such as the emerging standards for XML (see http://www.w3c.org), are still under development. Indeed it is hoped that, as OPAL is developed, it will help to shape these standards and technologies.

Further information can be found at:

• •

  The OPAL Project ­ http://oulib1.open.ac.uk/wh/research/opal/

• •

  Open Libr@ry ­ http://oulib1.open. ac.uk/

• •

  The Knowledge Media Institute ­ http://kmi.open.ac.uk/

David Bradbury(D.C.Bradbury@ open.ac.uk) is OPAL Research Fellow Georgina Payne(G.F.Payne @open.ac.uk ) is OPAL Research Assistant at the Open University.