Interrogation d’une base de données à partir d’une base de connaissances dans un environnement de développement de connaissances.

Abstract

Expert systems are essentially software programs that simulate the reasoning of an expert, so he draws his experience in a specific area. Many expert systems have been successfully implemented to solve real problems. What is a good sign in terms of AI, however, major difficulties encountered during the extraction of expert knowledge, however, their construction form may be a weakness difficult to avoid in expert systems. To lift these constraints were created using an expert system technique for extracting knowledge from data (KDD), to form an initial knowledge base of a share. And secondly we want to classify the attributes of objects at the inference of the expert system module called the classification of these knowledge bases extracted by a learning algorithm (EDC) of decision trees.