Task Centered Representation for Expert Systems at the Knowledge Level

This work comes within the framework of current research on the development of functional architectures for second generation expert systems. The aim is to propose a model that allows the elicitation of a problem solving process at a “more appropriate” level of abstraction. The TASK model being developed has been specified in such a way as to offer the possibility of expressing the specificity of the semantics and the control structures implicit in certain clearly recognized reasoning modules whatever the level of abstraction is (generic tasks relating to a high-level cognitive activity or more specific tasks involved in solving a particular problem). This model is based on the key concept of “task”, a data structure which makes it possible to clearly reflect all the conceptual operations involved in the type of reasoning to be modelled. TASK denotes at one and the same time a representation model and the general architecture supporting task based system.