This paper presents a Genetic Algorithm for Production Systems
Optimization (GAPSO). The GAPSO finds an ordering of Condition Elements
(CEs) in the rules of a Production System (PS) that results in a (near)
optimal PS with respect to execution time. Finding such an ordering can
be difficult since there is often a large number of ways to order CEs
in the rules of a PS. Additionally, existing heuristics to order CEs in
many cases conflict with each other. The GAPSO is applicable to PSs in
general and no assumptions are made about the matching algorithm or the
interpreter that executes the PS. The results of applying the GAPSO to
some example PSs are presented. In all examples, the GAPSO found an
optimal ordering of CEs in a small number of iterations.