Ripple down rules (RDR), that is rules with hierarchical exceptions, are used in knowledge acquisition because they provide a well intelligible and modifiable representation for even very large expert systems. In this paper a formal semantics for RDRs is proposed, that covers first order rules as well as attribute-value based rules. An algebraic foundation is proposed, including simplification of RDRs and transformation of RDRs into flat lists of rules and ripple down rule sets, hence these knowledge representation schemes are put into perspective. It is shown, that a RDR has a shorter description length than an equivalent decision list. Induction of rules with exceptions is characterized as bidirectional movement in the hypothesis space, while known algorithms for learning rules or decision trees either perform a top-down specialization of the most general or a bottom-up generalization of the most special hypothesis. Known algorithms for induction of RDRs are summarized and compared a...

: The problem statement and scope of SISYPHUS III does not draw attention to one of the major problems faced in knowledge engineering (KE), which is building systems based on multiple sources of expertise. In this circumstance, the KE task becomes a requirements engineering (RE) task. A problem with many RE approaches is that the cost of use is prohibitive, and therefore such approaches are rarely applied. We present an RE strategy designed to handle conflicting perspectives that is an extension to current KE techniques. We implement this approach in the context of formal concept analysis (FCA) and ripple-down-rules (RDR) and describe an instantiation using the SISYPHUS III data. Our evaluation technique shows that the resolution operators have reduced the degree of conflict between viewpoints. 1. Introduction The SISYPHUS III experiment offers an excellent example of the similarity between the needs of knowledge engineering (KE) using multiple sources of expertise and those of require...

Abstract: Ripple down rules have addressed two of the major limitations of first generation Expert Systems (ES), the maintenance and knowledge acquisition (KA) bottleneck problems. This is achieved through acquiring knowledge directly from an expert, the use of an exception structure for knowledge representation and the storing of the cornerstone case associated with each rule. Just as RDR has offered a paradigm shift in the way these problems were solved, it is expected that RDR can offer a new approach to the issue of knowledge reuse. Due the poor acceptance of ES by end-users, our focus is more on reusing knowledge in different modes, such as explanation, critiquing or ‘what-if ’ within the same domain rather than the more conventional approach of reusing problem-solving methods or ontologies to solve a similar problem in a somewhat differerent domain. An evaluation of RDR for reuse showed that many modes of use were possible without any change to the knowledge or its structure but that some modes required understanding of the models represented. Since RDR does not require analysis or modeling of the domain for KA, maintenance or finding conclusions we have incorporated ideas from Formal Concept Analysis (FCA) to allow concepts and the relationships between them to be identified and explored. The addition of FCA tools to RDR is described in this paper. 1. The Reuse of Knowledge The reuse of knowledge should result in potential savings

. The development of highly effective heuristics for search problems is a difficult and time-consuming task. We present a knowledge acquisition approach to incrementally model expert search processes. Though, experts do not normally have introspective access to that knowledge, their explanations of actual search considerations seems very valuable in constructing a knowledge level model of their search processes. The incremental method was inspired by the work on Ripple-Down Rules which allows knowledge acquisition and maintenance without analysis or a knowledge engineer. We substantially extend Ripple Down Rules to allow undefined terms in the conditions. These undefined terms in turn become defined by Ripple Down Rules. The resulting framework is called Nested Ripple Down Rules. Our system SmS1.2 (SmS for Smart Searcher), has been employed for the acquisition of expert chess knowledge for performing a highly pruned tree search. Our first experimental results in the chess domain are ev...

The work reported in this thesis is motivated by the belief that knowledge-based systems (KBS) research needs to focus more on users ’ needs and cater for the various decision situations in which users will find themselves. To build individual systems that cater for all the activities that may be needed is not feasible or desirable. The problems associated with capturing knowledge are well known and the ability to capture knowledge once and access and manipulate the knowledge in multiple ways is highly desirable. It adds value to the original knowledge and offers all the benefits associated with the reuse of resources. Thus, the problem becomes one of knowledge reuse. The research question pursued in this thesis is “can knowledge captured for one purpose, such as consultation, be reused to support a wide range of alternative purposes, such as critquing or tutoring, allowing the user to answer different types of questions according to their current circumstances”? Further, this question was to be answered in a situated cognition, dynamic knowledge framework. The system developed in this thesis is based on the Multiple Classification Ripple Down Rule (MCRDR) knowledge acquisition and representation technique. MCRDR is a form of casedbased

Abstract: The need for analysis and modeling of knowledge has been espoused by many researchers as a prerequisite to building knowledge based systems (KBS). This approach has done little to alleviate the knowledge acquisition (KA) bottleneck or the maintenance problems associated with large KBS. For actual KA and maintenance we prefer to use a technique, known as ripple down rules (RDR) that is simple, yet reliable, and later see what models can be produced from the knowledge for the purpose of reuse. Tools based on Formal Concept Analysis have been added to RDR to uncover and explore the underlying conceptual structures. 1 Models and their Role in Knowledge Acquisition Since Newell’s [21] paper on “The Knowledge Level ” there has been increasing awareness and acceptance of the need to model knowledge at a level above its symbolic representation. This notion was further explored by Clancey [3] who used task and problem solving methods analysis to divide problems into “heuristic classification ” and “heuristic construction”. Following Van de Velde [33] approaches

The major focus of recent knowledge acquisition research has been on problem-solving methods (PSM). This paper present results where a PSM developed for classification has been extended to handle a configuration or parametric design task, designing ion chromatography methods in analytical chemistry. Surprisingly good results have been obtained seemingly because any knowledge that has been added to the knowledge base, has been added precisely to overcome any limitations of the PSM. These results suggest a trade-off between domain knowledge and the power of the PSM and that greater use of domain knowledge would facilitate re-use by allowing PSMs to be used for a broader range of tasks. INTRODUCTION The critical insight with the emergence of knowledge based systems (KBS), was the usefulness of knowledge as compared to pure search. This was formulated in ideas such as the "knowledge principle". This resulted in expert system shells which while specifically intended for the accumulation o...

Knowledge-level (KL) modeling can be characterised as theory subset extraction where the extracted subset is consistent and relevant to some problem. Theory subset extraction is a synonym for Newell's principle of rationality, Clancey's model construction operators, and Breuker's components of expert solutions. In an abductive framework, a PSM is the extraction controller and is represented by a suite of BEST inference assessment operators. Each BEST operator is a single-classification expert system which accepts or culls a possible inference. PSMs can therefore be maintained by rippledown -rules, a technique for maintaining singleclassification expert systems. 1 Introduction Newell's knowledge-level (KL) approach modeled intelligence [37] as a search for appropriate operators that convert some current state to a goal state. Domain-specific knowledge are used to select the operators according to the principle of rationality; i.e. an intelligent agent will select an operator which i...

. Ripple down rules (RDR) is a knowledge acquisition technique that addresses the bottleneck problem by allowing rapid development of knowledge bases (KB) by experts, without the need for lengthy analysis or intervention of a knowledge engineer. This is achieved through the use of an exception structure and the storing of cornerstone cases. The exception structure avoids the problem of side-effects that occur when traditional rule-based ES are maintained, as knowledge in an RDR KB is never deleted or changed, only added. This can lead to repetitious knowledge. While studies have shown that the repetition problem is small, the concern of this study is the impact of repetition on reuse of the knowledge base for purposes such as explanation, modeling or tutoring. This paper reports on work that has been done using two different machine learning techniques, Induct and Rough Sets, to compact various ripple down rule knowledge bases by removing repetitious or redundant knowledge. 1. Introduc...

: This paper offers a framework and a process model for comparing requirements from different stakeholders. The framework combines a knowledge based system approach to requirements elicitation, a mathematically based technique for conceptual modelling, a four-state model of comparison to identify different types of conflict, a number of strategies to resolve inconsistencies and a distance metric for determining if viewpoints are converging. The framework is an extension of a knowledge engineering framework for reconciling differences between multiple sources of expertise. To justify this extension a comparison is made between the nature of knowledge and requirements and the issues that face both disciplines. The paper focuses on the common issues of maintenance and combining requirements, which are regarded as a specialised type of knowledge. Keywords: requirements engineering, knowledge engineering, ripple down rules, formal concept analysis 1 Using Knowledge Engineering Techniques ...