An Expert System for Evaluating the ‘Knowledge Potential’ of Databases

Potter, Dr Stephen and Aitken, Dr Stuart (2001) An Expert System for Evaluating the ‘Knowledge Potential’ of Databases. In Bramer, Prof M, Eds. Proceedings The Twenty-first SGES International Conference on Knowledge Based Systems and Applied Artificial Intelligence, Cambridge, UK.

This document is not available online.

Data Mining techniques can, under favourable conditions, extract valuable knowledge from an organisation's databases. However, the precise nature of these favourable conditions is poorly articulated, and as a result organisations run the risk of instigating costly and time-consuming data mining episodes upon inappropriate or irrelevant data. In response to this problem, the authors have applied expert systems technologies to the task of predicting the extent to which a given database contains knowledge that is both valuable and susceptible to mining. It is intended that this system will form a component of a `knowledge auditing' method for appraising an organisation's (existing and potential) knowledge resources. This paper describes this application, along with some of the particular issues surrounding its implementation, testing and delivery to prospective users.

Contact the site administrator at: hg@ecs.soton.ac.uk