Weak Equivalence for Constraint Sets

摘要

A generalization of equivalence between constraint sets, called weak equivalence,is described. This new equivalence relation takes into account that not all variables have the same function in a constraint set and therefore distinguishes between restriction variables and intermediate variables. The properties of weak equivalence and its underlying notion of weak implication with an axiomatic approach, are explored. In addition a complete set of axioms for weak implication is presented. With examples derived from the declarative rule language RL, the applicability of weak equivalence to constraint solving is derived.