Syntacticness, Cycle-Syntacticness and Shallow Theories (1994)
| Venue: | Information and Computation |
| Citations: | 9 - 0 self |
BibTeX
@ARTICLE{Comon94syntacticness,cycle-syntacticness,
author = {Hubert Comon and Marianne Haberstrau and Jean-pierre Jouannaud},
title = {Syntacticness, Cycle-Syntacticness and Shallow Theories},
journal = {Information and Computation},
year = {1994},
volume = {111},
pages = {154--191}
}
Years of Citing Articles
OpenURL
Abstract
Solving equations in the free algebra T (F; X) (i.e. unification) uses the two rules: f(~s) = f( ~ t) ! ~s = ~ t (decomposition) and s[x] = x !? (occur-check). These two rules are not correct in quotients of T (F; X) by a finitely generated congruence =E . Following C. Kirchner, we first define classes of equational theories (called syntactic and cycle syntactic respectively) for which it is possible to derive some rules replacing the two above ones. Then, we show that these abstract classes are relevant: all shallow theories, i.e. theories which can be generated by equations in which variables occur at depth at most one, are both syntactic and cycle syntactic. Moreover, the new set of unification rules is terminating, which proves that unification is decidable and finitary in shallow theories. We give still further extensions. If the set of equivalence classes is infinite, a problem which turns out to be decidable in shallow theories, then shallow theories fulfill Colmerauer's indep...







