There seems to be general agreement amongst those involved in KBS research that in order to be useful, a system must be able to explain its reasoning to a user. This paper reviews the development of explanation facilities in knowledge-based systems. It differentiates between explanation as a problem-solving process, and that which explains a reasoning process. This review concentrates on the latter, identifying and giving examples of three categories of reasoning explanation.

We then look at user requirements for explanation. What makes an explanation useful depends on the expectations of a user, which in turn depends on such issues as user background and system context. Several techniques are examined that have been applied to the problem of producing explanations that are appropriately structured and conveyed.

Finally, we discuss some of the work that has been done in describing theories of human discourse and explanation, and some issues that will become increasingly important for future explanation systems.

An extensive annotated bibliography is provided.

Computational modelling of scientific discovery is emerging as an important research field in artificial intelligence. Various computational systems modelling different aspects of scientific research and discovery have been developed. This paper looks at some of these models in order to examine how knowledge is organized in such systems, what forms of representation they have, how their methods of learning and representation are integrated, and the effects of representation on learning. The paper also describes the achievements and shortcomings of these systems, and discusses the obstacles in developing more comprehensive models.

Constraint logic programming (CLP) is a generalization of logic programming (LP) where unification, the basic operation of LP languages, is replaced by constraint handling in a constraint system. The resulting languages combine the advantages of LP (declarative semantics, nondeterminism, relational form) with the efficiency of constraint-solving algorithms. For some classes of combinatorial search problems, they shorten the development time significantly while preserving most of the efficiency of imperative languages. This paper surveys this new class of programming languages from their underlying theory, to their constraint systems, and to their applications to combinatorial problems.

Intelligent tutoring systems can be seen as a next step for computer-based training systems, but also as an important by-product of knowledge-based expert systems. This paper surveys the development and progress in the area, with a special emphasis on the potential for an emerging engineering discipline as opposed to a mere crafting of systems. Major components in intelligent tutoring systems as realized so far are discussed, and key issues for successful future development identified. Knowledge representation, student modelling, planning, natural language issues, explanations and learning are discussed in more depth as being the cornerstones of both tutoring systems and artificial intelligence. Examples from specific implementations are used to illustrate key points. In the concluding discussion we present our attempt at dealing with some of the problems facing the area. In the project Knowledge-Linker, we aim at extending the functionality of a knowledge-based system with tutoring capabilities, and suggest one way of explicitly dealing with teaching strategies.

This paper presents a survey of machine induction, studied mainly from the field of artificial intelligence, but also from the fields of pattern recognition and cognitive psychology. The paper consists of two parts: Part I discusses the basic principles and features of the machine induction process; Part II uses these principles and features to review and criticize the major supervised attribute-based induction methods. Attribute-based induction has been chosen because it is the most commonly used inductive approach in the development of expert systems and pattern recognition models.

Expert systems are being built despite the widely acknowledged problem of acquiring knowledge from experts. This study attempts to understand how knowledge acquisition is conducted in practice by investigating three expert system development projects. A CASE research methodology is adopted, and data is collected through unobtrusive observation, from taped protocols of knowledge acquisition sessions, retrospective interviews with the participants involved, and deliverables produced. The variables examined include the problem domain, the domain expert, the knowledge engineer, the knowledge acquisition process, the expert system construction process, potential users, organizational setting, and the expert system itself. The knowledge acquisition processes for three expert systems in the domains of law of negligence, telephone line fault diagnosis, and wastewater treatment have been examined. By juxtaposing the observations drawn with findings from the relevant literature, the study makes prescriptive suggestions on considerations and techniques for future acquisition efforts, and provides data for hypothesis generation in further research.

This paper gives a state-of-the-art overview of the rapidly expanding field of user modelling in artificially intelligent systems. After showing how a user modelling component can improve parts of the processing of, and the interaction with, a large number of systems, the current situation in this field is sketched by means of a number of short descriptions of systems employing user modelling techniques. Next, a synthesis of this review is made by discussing the aspects on which the existing methods differ. It turns out that these aspects can best be approached from two points of view: the technical level and a more abstract, functional level. This dichotomy results in eight dimensions on which to compare the modelling methods. The synthesis is then completed by describing the existing modelling techniques, and classifying the reviewed systems. In the final section, current trends in the field are outlined and prerequisites for acceptance of user modelling on a larger scale are discussed.

Learning is one of the important research fields in artificial intelligence. This paper begins with an outline of the definitions of learning and intelligence, followed by a discussion of the aims of machine learning as an emerging science, and an historical outline of machine learning. The paper then examines the elements and various classifications of learning, and then introduces a new classification of learning based on the levels of representation and learning as knowledge-, symboland device-level learning. Similarity- and explanation-based generalization and conceptual clustering are described as knowledge level learning methods. Learning in classifiers, genetic algorithms and classifier systems are described as symbol level learning, and neural networks are described as device level systems. In accordance with this classification, methods of learning are described in terms of inputs, learning algorithms or devices, and outputs. Then there follows a discussion on the relationships between knowledge representation and learning, and a discussion on the limits of learning in knowledge systems. The paper concludes with a summary of the results drawn from this review.

Many problem domains exhibit inherent parallelism, and parallel systems which capture and exploit this can be used to look for efficient solutions to AI problems. Functional programming languages are expected to be efficiently realisable on fifth generation hardware. A rational reconstruction of AI programming paradigms is used to investigate the programmability and performance of functional languages in this particular area.

Three languages—Standard ML, Hope+ and Miranda—are used in the rational reconstruction, each language being used to implement three applications. Results indicate that functional programming languages have become much more useable in recent years, and have the potential to become useful tools in AI problem solving. A brief annotated bibliography of texts which covers the introduction to, theory and implementation issues of, functional programming languages, is included.

In the coming years, scheduling will be an increasingly important activity in the manufacturing and aerospace industries, especially at the US National Aeronautics and Space Administration (NASA). Expert systems have successfully been used to aid in the scheduling function. This paper surveys some of the conflict resolution approaches used in building expert scheduling systems, and then examines the feasibility of using an expert systems approach for scheduling activities and resolving conflicts for experimenters to use NASA Goddard-supported satellites. Recommendations for further reading in expert scheduling systems are given.