KB refinement: a new intercative method

This paper presents a new approach of interactive knowledge base refinement which guides the expert in the modification of the `faulty´ knowledge. The first step of our method is a validation stage which checks the consistency and the completeness of the knowledge base. Then, the second step is a refinement stage which aims to eliminate the anomalies that have been detected. Instead of proposing different sets of modifications of the knowledge base, our approach detects elements of the knowledge which are easily understandable by the expert and propose modifications on these elements. We propose to build a set of refinements of the knowledge base by using an iterative process in a close interaction with the expert. We present to the expert an understandable element of the knowledge base that is involved in faulty behaviour and a set of modifications on these elements that contribute to solve these faults. The expert makes a choice between the different modifications, then the updating is made according to this choice.