Systematic Building of Conceptual Classification Systems with C-KAT

C-KAT is a method and a tool which supports the design of “feature oriented” classification systems. During the design of these systems, one is very often confronted with the problem of the “calculation of the attribute cross-product”. It arises because the examination of the dependency and compatibility relations between the attributes leads to the need to generate the cross-product of their features. The C-KAT method uses a specialised Heuristic Classification conceptual model named “classification by structural shift” which sees the classification process as the matching of different classifications of the same set of objects or situations organised around different structural principles. To manage the complexity induced by the cross-product, C-KAT supports the use of a least commitment strategy which applies in a context of constraint-directed reasoning. The method is presented using a detailed example from the field of industrial fire-insurance.