Authors: Kamal Kansou, Guy Della Valle, Amadou Ndiaye
We present a procedure of knowledge representation based on a qualitative algebra, to predict the wheat flour dough behaviour from mixing settings. The procedure guarantees the consistency of the knowledge base and provides a concise and explicit representation of the knowledge. The qualitative model is implemented as a knowledge-based system (KBS) accessible and understandable by scientists and technologists in breadmaking. The KBS is a record of the domain knowledge, mainly know-how, and a tool to confront predictions of the dough condition with real observations. An example of such a confrontation about the wheat flour dough mixing process is shown; the results gives insight into ill-known relations between the process settings and the dough condition.