Selecting and Generating Concept Structures

Abstract

Knowledge can be used to build applications that carry out a number of complex activities (e.g., diagnosis, monitoring, decision-making). However, knowledge is difficult to elicit and not easy to represent. Moreover, the way knowledge may be related to other forms of data (e.g., collected in human-computer interaction) such that computer-based problem-solving becomes possible, is often unclear. In this paper, an approach to knowledge management is described that uses sequences of symbols. This sort of data is a by-product of numerous human activities, but it is also the output of many process control devices. We present some examples that support the claim that this method lessens the burden of knowledge acquisition and opens new perspectives for knowledge management.