Reference: Egar, J. W. Graph-Grammar Assistance for Knowledge Acquisition. Knowledge Systems Laboratory, April, 1992.
Abstract: One of the most difficult aspects of modeling complex medical dilemmas in decision-analytic terms is composing a diagram of relevance relations from a set of domain concepts. The concepts, as presented to the knowledge engineer, generally have no inherent order, yet they have charcteristic positions and typical roles in a semantic network model. We have been using a graph-grammar production system to express such inherent interrelationships among medical terms. We have found that this graph-grammar system facilitates the modeling of medical dilemmas. We also suggest a translate-and-assemble perspective for certain knowledge-acquisition problems. We suspect that graph grammars,and the translate-and-assemble perspective, may be useful guides to modeling in many other circumscribed domains.