In qualitative reasoning research, much effort has been spent on developing representation and reasoning formalisms. Only recently, the process of constructing models in terms of these formalisms has been recognised as an important research topic of its own. Approaches addressing this topic are examined in this review. For this purpose a general model of the task of constructing qualitative models is developed that serves as a frame of reference in considering these approaches. Two categories of approaches are identified: model composition and model induction approaches. The former compose a model from predefined partial models and the latter infer a model from behavioural data. Similarities and differences between the approaches are discussed using the general task model as a reference. It appears that the majority of approaches focus on automating model construction entirely. Assessing and debugging a model in cooperation with a modeller is identified as an important topic for future research

This paper reviews a number of applications of machine learning to industrial control problems. We take the point of view of trying to automatically build rule-based reactive systems for tasks that, if performed by humans, would require a high degree of skill, yet are generally performed without thinking. Such skills are said to be sub-cognitive. While this might seem restrictive, most human skill is executed subconsciously and only becomes conscious when an unfamiliar circumstance is encountered. This kind of skill lends itself well to representation by a reactive system, that is, one that does not create a detailed model of the world, but rather, attempts to match percepts with actions in a very direct manner.

In recent years a number of programming languages based upon the direct execution of temporal logic formulae have been developed. The use of such logics provides a powerful basis for the representation and implementation of a range of dynamic behaviours. Though many of these languages are still experimental, they are beginning to be applied, not only in computer science and AI, but also in less obvious areas such as user interfaces, process control and social modelling. This article provides an introduction to some of the basic concepts of executable temporal logics, together with an overview of the main approaches being pursued.

This article is an essay on directions and methodology in computer-science oriented research on scientific discovery. The essay starts by reviewing briefly some of the history of computing in scientific reasoning, and some of the results and impact that have been achieved. The remainder analyses some of the goals of this field, its relations with sister fields, and the practical applications of this analysis to evaluating research quality, reviewing, and methodology. An earlier review in this journal (Kocabas 1991b) analysed scientific discovery programs in terms of their designs, achievements and shortcomings; the focus here is research directions, evaluation and methodology, all from the viewpoint of computer science.

In his thought-provoking paper, Valdes-Perez (1996, this volume) carefully describes the methodology and future directions for research in Machine Scientific Discovery (MSD), as viewed predominantly from the standpoint of computer science. Here, I shall offer some remarks from the angle of domain sciences (despite the fact that my basic area of expertise is actually general and computational linguistics).

Very little in computing these days is promoted with as much vigour as multimedia. Multimedia, we are told, will increase the usability and productivity of systems; it will improve the ability of people to learn from educational applications; it will combine naturally with other interface technologies to create a more natural interaction than ever before.