Item

Abstract

In an ICALP (1991) paper, H. Chen and J. Hsiang introduced a notion that allows for a finite representation of certain infinite sets of terms. These so called w-terms find an application in logic programming, where they can serve to represent finitely an infinite number of answers or to avoid nontermination in certain cases. Another application is in the field of equational logic. Using w-terms, it is possible to avoid a certain type of divergence of ordered completion. In all cases, unification is the basic computational aspect of this notation. Chen and Hsiang give a complete and terminating unification algorithm for w-terms. Recently, H. Comon introduced terms with exponents, thus significantly extending Chen and Hsiang's notion of w-terms. He provides a fairly complicated unification algorithm. This paper introduces a further syntactic generalization of Comon's notion together with a comparatively simple inference system for unification.