Phillip Preston

Knowledge Acquisition without Analysis

This paper suggests that a distinction between knowledge acquisition methods should be made. On the one hand there are methods which aim to help the expert and knowledge engineer analyse what knowledge is involved in solving a particular type of problem and how this problem solving is carried out. These methods are concerned with classifying the different types of problem solving and providing tools and methods to help the knowledge engineer identify the appropriate approach and ensure nothing is omitted. A different approach to knowledge acquisition focuses on ensuring incremental addition of validated knowledge as mistakes are discovered (validated knowledge here means only that the earlier performance of the system is not degraded by the addition of new knowledge). The organisation of this knowledge is managed by the system rather than the expert and knowledge engineer. This would seem to correspond to human incremental development of expertise. From this perspective task analysis is a secondary activity related to explanation and justification not acknowledge acquisition. Ripple Down Rules is a limited example of this approach. The paper considers the possibility of extending this approach to make it more generally applicable.

Local Patching Produces Compact Knowledge Bases

Knowledge acquisition (KA) encompasses working with the expert to model the domain and a suitable problem solving method as preconditions for building a knowledge based system (KBS) and secondly working with the expert to populate the knowledge base. Ripple Down Rules (RDR) focuses on the second of these activities and allows an expert to populate a knowledge base (KB) without any knowledge engineering assistance. It is based on the idea that since the knowledge an expert provides is a justification of his or her judgment given in a specific context, this knowledge should only be used in the same context. Although the approach has been used for large single classification systems, it has the potential problem that the local nature of the knowledge may result in much repeated knowledge in the KB and much repeated knowledge acquisition. The study here attempts to quantitate and compare KB size and performance for systems built by experts with various levels of expertise and also inductively. The study also proposes a novel way of conducting such studies in that the different levels of expertise were achieved by using simulated experts. The conclusion from this study is that experts are likely to produce reasonably compact and efficient knowledge bases using the Ripple-Down Rule approach.

Teaching a computer ion chromatography from a database of published methods

As ion chromatography (IC) has matured as an analytical technique it has become more automated. Most instrument control and data handling is now handled by computers. However, IC has not seen the abundance of automated method optimisation techniques which are provided to conventional chromatography. To a certain extent this was because IC differed greatly in the approach required to optimise selectivity and sensitivity. There was quite a diverse range of chemistries (or separation mechanisms) applicable to IC, such as ion exchange, ion interaction, etc. This paper describes an effort to fill this gap by developing an expert system which can give comprehensive advise on suitable method conditions for a variety of IC mechanisms. To build this system we applied an approach known as induction by machine learning, which was developed within the field of artificial intelligence (AI). A database of over 4000 published methods using IC, where the sample information and the chromatographic conditions were recorded, was used to train an expert system (ES). Both induction and a neural network model were applied to this task and an expert system which can advise on the following IC method conditions: mobile phase, column, pH, mechanism, post-column reactors, suppressor use and gradient applicability, was successfully developed. This paper presents a summary of the most pertinent conclusions from this study. A test set of different methods was extracted from the database and they were not applied in the training of the expert system. These were used to test the expert system and different amounts of information were used as inputs. The resulting outputs of the expert system were evaluated by the expert, who decided whether the method would work or not and if it was a good method or the ideal method for the application. Over 85% of methods were found to work and almost 62% of the methods were considered ideal. These were acceptable results when one considers the limitations of using a database of published methods as a learning set and the time saved by the use of machine learning.

Towards an expert system in ion-exclusion chromatography by means of multiple classification ripple-down rules

We describe the development and maintenance of an expert system to advise on the configuration of systems for ion-exclusion chromatography. The aim of the system is to define appropriate conditions for the separation of desired groups of acids or bases. The system is implemented in a rule-based system, multiple classification ripple-down rules, which offers multiple conclusions from rules based on the attributes of the system. In this case the attributes include physical and chemical properties of the solutes and the availability of instrumentation and accessories. With this information the method conditions can be defined for the detector, mobile-phase, whether suppression is to be used, and other ion-exclusion chromatography method conditions. A unique feature is that some conditions may be filled in by the program or be given by the user. Because of the nature of the “ripple-down rules” approach, in which new knowledge is always added as an amendment to an existing conclusion (and therefore cannot interfere with other conclusions), the expert or user can maintain and alter the system easily according to their own needs without the help of a software engineer. The system was developed and tested using cases from published papers on ion exclusion chromatography. For a set of 83 cases, although the expert system only agreed with the published conditions in 53% of cases, when the predictions were assessed by a recognized ion chromatography expert, 88% were pronounced “workable”.