Anti-logicist framework for design-knowledge representation - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Annual Reviews in Control Année : 2015

Anti-logicist framework for design-knowledge representation

Résumé

The knowledge reuse and mapping are among the most important concerns related to the design knowledge representation. In this paper, authors focus on the importance of one specific property of a design knowledge representation: the unambiguity. Authors show 1) how the ambiguity of the representation can increase the risk of a failure in the reuse and mapping processes, 2) how most of works in the literature use formal logic constructs and finally 3) how the use of these can increase the risk of ambiguity. On the basis of these remarks, an overview on the works on the anti-logicist architecture is provided: the systems based on this architecture show an intelligent behaviour without using logic constructs. An analysis and a transposition of the anti-logicist principles are then performed to build a framework allowing to represent design knowledge without logic constructs. To do so 1) main concepts are formalised in a conceptual model; 2) an algorithm has been designed to map pieces of knowledge based only on the representation syntax; 3) two instantiations of the framework are showed using a CAD instantiation. Finally, the limits of the current deployment of the framework and the research perspectives are discussed.
Fichier principal
Vignette du fichier
Giovannini A. - ARC.pdf (2.36 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01122807 , version 1 (04-03-2015)

Identifiants

Citer

Antonio Giovannini, Alexis Aubry, Hervé Panetto, Hind Bril El-Haouzi, Ludovic Pierrel, et al.. Anti-logicist framework for design-knowledge representation. Annual Reviews in Control, 2015, 39 (1), pp.144-157. ⟨10.1016/j.arcontrol.2015.03.013⟩. ⟨hal-01122807⟩
183 Consultations
211 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More