Premium
Toward a theory of verification and validation: Artifacts
Author(s) -
O'Leary Daniel E.
Publication year - 1994
Publication title -
international journal of intelligent systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.291
H-Index - 87
eISSN - 1098-111X
pISSN - 0884-8173
DOI - 10.1002/int.4550090906
Subject(s) - computer science , artifact (error) , set (abstract data type) , verification and validation of computer simulation models , artificial intelligence , functional verification , quality (philosophy) , verification and validation , formal verification , programming language , mathematics , philosophy , statistics , epistemology
This article uses the notion of artifacts to generate a theory for the verification and validation of intelligent systems. This theory has a number of applications. First, it provides a language and a way of thinking about verification and validation. Second, it suggests that we examine not the processes used in the design and development of systems, but instead the tangible artifacts generated by those processes. Third, it broadens the base of the set of artifacts encompassed by the typical verification and validation processes to include more than just specifications and prototypes. Fourth, the use of artifacts leads to a redefinition of verification and validation in terms of artifacts. Fifth, verification and validation tests are tied to artifacts, so that the choice of a set of artifacts can be used as a means of generating a portfolio of tests to investigate each artifact and the interaction of artifacts. Sixth, relationships between artifacts can be defined as “operators” which can be isolated and investigated for their individual quality. Seventh, standards can be generated for specific artifacts and operators rather than for entire development processes. © 1994 John Wiley & Sons, Inc.