Fixpoint Techniques for Non-Monotone Maps

Using a technique of successive approximations, conditions were derived under which such approximations converge to a fixpoint. In such cases, the limit of a sequence of approximations provides a mathematically computable fixpoint for non-monotone maps. In this paper, we define a class of logic programs for which such a technique provides a way of constructing a minimal fixpoint with special properties—stable model, preferable to any other minimal fixpoint. Unlike stratified programs, the class of programs we present has the advantage of allowing negation through recursion. We also present some results on unique fixpoints in non-monotone maps.