Knowledge acquisition by machine learning and data mining

Abstract

A critical problem in the development of knowledge-based systems is capturing knowledge from the experts. There are many knowledge elicitation techniques that might aid this process, but the fundamental problem remains: tacit knowledge that is normally implicit, inside the expert's head, must be externalised and made explicit. Knowledge acquisition (KA) thus has been well recognised as a bottleneck in the development of knowledgebased systems and is a key issue in knowledge engineering. Traditionally, KA techniques can be grouped into three categories: manual, semi-automated (interactive) and automated (machine learning (ML) and data mining). Since the early days of artificial intelligence (AI), the problem of KA, the elicitation of expert knowledge in building knowledge bases, has been recognised as a fundamental issue in knowledge engineering. © 2007 Springer-Verlag Berlin Heidelberg.