Abstraction of Objects by Conceptual Clustering

Very bound to the logic of first-rate predicates, the formalism of conceptual graphs constitutes a knowledge representation language. The abstraction of systems presents several advantages. It helps to render complex systems more understandable, thus facilitating their analysis and their conception. Our approach of conceptual graphs abstraction, or conceptual clustering, is based on rectangular decomposition. It produces a set of clusters representing similarities between subsets of objects to be abstracted, organized into a hierarchy of classes: the Knowledge Space. Some conceptual clustering methods already exist. Our approach is distinguishable from other approaches in as far as it allows a gain in space and time.