Modelling Uncertainty in Expertise

Almost all approaches of model-based development of knowledge-based systems are lacking an explicit handling of uncertain knowledge. Based on a KADS-oriented model of expertise our paper presents a general framework that enables the representation of uncertainty in a structure of causality. Based on an existing model of expertise we introduce a separate model of uncertainty, whose elements are represented independent of speci c numerical processing methods. With this approach we got the foundations of a model-based framework integrating di erent kinds of uncertain information. The corresponding elements of both models are explicitly connected. The results of our work give prospects for a new understanding of knowledge engineering considering uncertainty in all steps of system development.