The process of extracting, structuring and organizing elicited knowled
ge (called knowledge acquisition) is a bottleneck in developing knowle
dge-based systems. A manual approach that elicits domain knowledge by
interviewing human experts typically has problems, because the experts
are often unable to articulate their reasoning rules. An automatic ap
proach that induces knowledge from a set of training cases also suffer
s from the unavailability of sufficient training cases. We present an
integrated approach that combines the strengths of both methods to com
pensate for their weaknesses. In this approach, human experts are resp
onsible for solving problems, whereas an inductive learning algorithm
is responsible for reasoning and consistency checking. Copyright (C) 1
996 Elsevier Science Ltd