Qualitative reasoning

In the mid-1980s, AI turned its attention to qualitative reasoning (QR) as a means of constructing representations that provide a better understanding of the physical phenomena involved in a process. The development of techniques for qualitatively representing the world in order to "schematically understand how it works" is a very natural and consequently very old idea. Since supervision relies more on the notion of relevance than on the accuracy of representations, QR quickly found a privileged field of application.

Above all, qualitative modeling must be simpler than classical physics. The simplification method commonly considered is to reduce the precision of numerical modeling; taken to the extreme by De Kleer [13]...