A knowledge-level testing method

This paper aims at describing a testing method devoted to the study of knowledge-based systems (KBS) behaviour and results. Like any validation method, this method relies on specifications. Moreover, as the focus of the study is KBS behaviour, these specifications need to be expressed at high level. We distinguish between two kinds of specifications, system specifications and anomaly specifications, depending on the validation objectives. We adopt a knowledge-level perspective of validation since our first objective is to compare the KBS behaviour to expected behaviours that are a part of system specifications and described within a KADS conceptual model. In order to detect inconsistent results, we propose to add some kind of validation knowledge to the domain level knowledge. This method is illustrated by the validation of a prototype for natural language understanding.

[1]  Corinne Haouche.  Utilisation des concepts de l'acquisition des connaissances pour valider un SBC , 1995 .

[2]  Corinne Haouche.  Testing a KBS Using a Conceptual Model , 1994, AAAI.

[3]  Pedro Meseguer,et al.  Verification of Multi-Level Rule-Based Expert Systems , 1991, AAAI.

[4]  Marc Ayel,et al.  SACCO-SYCOJET: two different ways of verifying knowledge-based systems , 1991 .

[5]  Bob J. Wielinga,et al.  KADS: a modelling approach to knowledge engineering , 1992 .

[6]  Jean-Paul Krivine,et al.  L'acquisition des connaissances vue comme un processus de modélisation : méthodes et outils , 1991 .

[7]  Nathalie Aussenac-Gilles,et al.  Making a method of problem solving explicit with MACAO , 1994, Int. J. Hum. Comput. Stud..

[8]  Allen Newell,et al.  The Knowledge Level , 1989, Artif. Intell..

[9]  Marie-Christine Rousset,et al.  On the consistency of knowledge bases: the COVADIS system , 1988, Comput. Intell..

[10]  Walton A. Perkins,et al.  Checking an Expert Systems Knowledge Base for Consistency and Completeness , 1985, IJCAI.

[11]  Bob J. Wielinga,et al.  Using explicit ontologies in KBS development , 1997, Int. J. Hum. Comput. Stud..

[12]  Jean-Pierre H. Laurent.  Proposals for a Valid Terminology in KBS Validation , 1992, ECAI.

[13]  Jean Charlet,et al.  Causal model-based knowledge acquisition tools: discussion of experiments , 1996, Int. J. Hum. Comput. Stud..

[14]  Uma G. Gupta.  Validating and Verifying Knowledge-Based Systems , 1991 .

[15]  Jacques Bouaud,et al.  TREE: the Heuristic Driven Join Strategy of a RETE-Like Matcher , 1993, IJCAI.