Heuristics for ordering of production rules
Srinivas Krovvidy, William G. Wee · Proceedings of the 17th conference on ACM Annual Computer Science Conference · 1989
In real time expert systems, the primary requirement is that the time taken in deciding a rule must be as small as possible. In this paper, we define the complexity of inference in terms of the number of conditions tested during the inference process. We present heuristics to arrange rules in an existing knowledge base (KB) to minimize this complexity. We also present heuristics to generate the rules automatically from learning system to minimize the complexity.