Quality assessment of knowledge-based software: some certification considerations

Abstract

Knowledge-based technology is being applied to complex problem-solving and critical tasks in many application domains. Concerns have naturally arisen as to the dependability of knowledge-base systems (KBSs). As with any software, attention to quality and safety must be paid throughout the development of a KBS, and rigorous verification and validation (V&V) techniques must be employed. Research in KBS V&V has emerged as a distinct field only in the last decade and is intended to address issues associated with KBS quality aspects and to credit such applications with the same degree of dependability as conventional applications. However, a major drawback of the current state of acceptance of KBS software is the lack of standardization. Although standards do exist for software, extensions are needed to cope with specific KBS software characteristics. This paper advocates the idea that, although quality assessment of KBS software is still a challenging issue, KBS software is not an impossible one. Some V&V techniques based on specific KBS software properties are described in this paper which aims at indicating KBS acceptability requirements may be specified and evaluated, being at the same time a good basis for certification.