Preservation of stronger equivalence in unfold/fold logic program transformation

Abstract

This paper shows that Tamaki-Sato's unfold/fold transformation of Prolog programs preserves equivalence in a stronger sense that of the usual least Herbrand model semantics, which Tamaki and Sato originally showed. Conventionally, the semantics of Prolog program is defined by the least Herbrand model. However, the least Herbrand model does not always characterize what answer substitutions are returned. This paper proves that any program obtained from an initial program by applying Tamaki-Sato's transformation returns the same answer substitutions as the initial program for any given top-level goal.