Integrated approach to the elicitation of reliable expert knowledge

In the basis of implicit learning systems, that is the systems to teach diagnostics skills, there is expert knowledge base. Since expert knowledge can rarely be verbalized by an expert one have to use special methods for the knowledge elicitation. Besides all the knowledge in the base should be absolutely reliable because it is designated to be put at the root of the learning system. This article proposes integrated approach to the elicitation of the reliable expert knowledge using ordinal classification, isolation of class borders, revelation of decision rules and zones of unstable expert knowledge.