Based on a survey of recent literature, this report aims to highlight the issues associated with the verification and validation of knowledge based systems. The confusion arising from the lack of clear terminology is considered, along with some of the characteristics of knowledge based systems that cause particular difficulties for verification and validation. The various approaches that can be adopted to address these difficulties are discussed, followed by a survey of recent research initiatives.

The author concludes that many of the difficulties associated with the verification and validation of knowledge based systems are a feature of the complexity of the system being built and the manner of its development rather than of the specific technology chosen to implement it.

This paper surveys the verification of expert system knowledge bases by detecting anomalies. Such anomalies are highly indicative of errors in the knowledge base. The paper is in two parts. The first part describes four types of anomaly: redundancy, ambivalence, circularity, and deficiency. We consider rule bases which are based on first-order logic, and explain the anomalies in terms of the syntax and semantics of logic. The second part presents a review of five programs which have been built to detect various subsets of the anomalies. The four anomalies provide a framework for comparing the capabilities of the five tools, and we highlight the strengths and weaknesses of each approach. This paper therefore provides not only a set of underlying principles for performing knowledge base verification through anomaly detection, but also a survey of the state-of-the-art in building practical tools for carrying out such verification. The reader of this paper is expected to be familiar with first-order logic.

The preceding paper on the refinement of rule-based systems by Precce et al. reflects a recent, rapid growth of interest in verification and validation of knowledge based systems. Much of the literature on this subject is thinly spread over conference proceedings, professional periodicals and academic journals, so we have tried to collect together some of the more frequently cited work to assist readers interested in the area.

It is usually agreed that programming languages for implementing (other) programming languages, or ‘implementation languages’, should be simple low-level languages which are close to the machine code and to the operating system. In this paper it is argued that a very high level implementation language is a good idea, of particular importance for knowledge-based systems, and that Lisp (as a language and as a system) is very well suited to be a very high level implementation language. The significance of special-purpose programming languages is also discussed, and the requirements that they have for a very high level implementation language are considered.

Abstract interpretation is a principled approach to inferring properties of a program's execution by simulating that execution using an interpreter which computes over some abstraction of the program's usual, concrete domain, and which collects the information of interest during the execution. Abstract interpretation has been used as the basis of research in logic and functional programming, particularly in applications concerned with compiler optimizations. However, abstract interpretation has the potential to be used in other applications, such as debugging or verification of programs. In this paper we review the use of abstract interpretation to both compiler optimizations and to other applications, attempting to give a flavour of the kind of information it is possible to infer and some of the issues involved