Expert Systems: The United Kingdom's Educational Approach
Why are expert systems (ES), which have so often failed to come up with the goods in practice, still considered to be important in academic curricula? More significantly, perhaps, why are schools of library and information science (LIS) taking the teaching of expert systems on board when the subject still has to prove itself in the commercial field? One of the reasons is that, although the current achievements of expert systems are modest, their potential is enormous.
Their development could be described as a small step for computing but a conceptual leap for mankind. Although, therefore, the commercially available systems and shells can appear singularly disappointing to the uninitiated, the intellectual achievement of persuading a machine to accept, manipulate, and reason with symbols and conduct a semi-intelligent dialogue with a user is considerable, and represents a breakthrough in the practical application of artificial intelligence.
Although expert systems are currently in their infancy, they are expected eventually to play a part in all future information systems, and this is often reason enough for LIS schools to feel that they should be taught in their curriculum. However, there are also other compelling reasons for their inclusion; LIS schools are concerned not only with exploring the potential of expert systems in LIS work, but also with exploring the potential of LIS professionalis in ES work.
In an ES, the practical knowledge and expertise of skilled human practitioners in specific domains is incorporated into the computer. This knowledge has to be collected -- from human or textual sources -- collated, refined, organized, represented, controlled, and retrieved. This is the bread-and-butter work of the LIS professional; many LIS schools, therefore, see exciting opportunities for LIS professionals to use their skills imaginatively in a new environment.
Expert systems, therefore, are a challenge to the LIS profession. How is this challenge being met by LIS schools? The authors have just undertaken a survey of ES teaching in the United Kingdom's seventeen LIS schools, under the auspices of a British Library-funded project investigating possible roles for information scientists in the development of expert systems. This article outlines some of the issues raised during discussions with the LIS teachers, and considers the implications for future graduates of LIS schools.
Teaching of Expert Systems in
U.K. Library Schools
All U.K. LIS schools teach ES work in some form, although the extent and depth of the teaching varies considerably from course to course. All include the subject as an integral part of other courses, such as information processing, information retrieval, information technology or library automation, but several also run entire modules concentrating solely on ES technology. These in-depth modules look at ES work in more detail, and students are usually expected to produce a small system, using an ES shell, as part of their assessed work.
Because of the time constraints imposed by a one-year postgraduate course, this intensive ES teaching is concentrated in the undergraduate courses. In all, half the undergraduate, and just under a quarter of the postgraduate courses, offer indepth ES options. The integrated courses, on the other hand, are taught to all students and are usually compulsory; the actual input varies from a mention in one lecture to several hours of intensive tutoring. The teachers of these courses felt that it was important for students to see the ES as a useful tool, to be used in several different contexts, rather than as an entirely separate issue.
Many of the staff spoken to felt that they had a comparatively free hand in deciding what they should include in the ES component of their courses, and a great deal depended upon the dynamism and enthusiasm of particular members of staff "sponsoring" the technology. …