Much of the work on validation and verification of knowledge based systems (KBSs) has been done in terms of implementation languages (mostly rule-based languages) .Recent papers have argued that it is advantageous to do validation and verification in terms of a more abstract and formal specification of the system . However , constructing such formal specifications is a dif ficult task . This paper proposes the use of formal specification languages for KBSdevelopment that are closely based on the structure of informal knowledge-models .The use of such formal languages has as advantages that (i) we can give strong support for the construction of a formal specification , namely on the basis of the informal description of the system ; and (ii) we can use the structural correspondence to verify that the formal specification does indeed capture the informally stated requirements .÷ 1996 Academic Press Limited
. IntroductionMuch of the current research in validation and verification (V&V) of knowledgebased systems (KBS) is done in terms of implementation languages , and mostly rule-based languages (Preece , Shinghal and Batarekh , 1992) . State of the art V&V systems check for such properties as loop-freeness of rule-bases , redundancy and reachability of rules , consistency etc . However , it is well known from software engineering , and it is argued in more recent work on V&V of KBS (Vermesan and Wergeland , 1994 a , b ) , that advantages can be gained by doing V&V in terms of a more abstract formal specification language , instead of an implementation language . If we are using an implementation-independent formal specification language , we can (i) do V&V in a much earlier stage of the life-cycle : we can already verify properties of the system in terms of its specification , before the expensive ef fort of making an implementation , and (ii) we can separate the verification of functional , implementation-independent properties from the verification of implementationspecific properties . Finally , (iii) V&V is facilitated by the higher level of mathematical abstraction that can be employed in an implementation-independent specification language . In order to obtain these advantages , we must overcome the dif ficulties † Present address :