The system we have developed, SYCOJET, automatically builds test cases making use of specific expert knowledge. In our first version, it uses the internal structure of the knowledge base to be tested, and implicitly assumes that the inference engine derives all that can be derived from the knowledge base and the problem data. Here we are concerned with how to take into account, in a second version of SYCOJET, the actual inference engine control strategy. This control strategy, included in the inference engine, is not in general explicited in a form accessible to the computer. In this paper, we investigate to what extent the knowledge of the control strategy of the system being tested can be used to improve the "quality" of the test cases generated by SYCOJET.
[1]
Alun Preece,et al.
State of the art in automated validation of knowledge-based systems☆
,
1994
.
[2]
Uma G. Gupta.
Validating and Verifying Knowledge-Based Systems
,
1991
.
[3]
Larry J. Morell.
Use of metaknowledge in the verification of knowledge-based systems
,
1988,
IEA/AIE '88.
[4]
Jean-Pierre H. Laurent.
Proposals for a Valid Terminology in KBS Validation
,
1992,
ECAI.
[5]
Jean-Pierre H. Laurent.
Control structures in expert systems
,
1984
.
[6]
Marc Ayel,et al.
SYCOJET and SACCO, two tools for verifying expert systems
,
1993
.
[7]
Johan de Kleer,et al.
An Assumption-Based TMS
,
1987,
Artif. Intell..