Foundations of logic programming; (2nd extended ed.)
Foundations of logic programming; (2nd extended ed.)
Alternating automata on infinite trees
Theoretical Computer Science
A finite presentation theorem for approximating logic programs
POPL '90 Proceedings of the 17th ACM SIGPLAN-SIGACT symposium on Principles of programming languages
Tree automata, Mu-Calculus and determinacy
SFCS '91 Proceedings of the 32nd annual symposium on Foundations of computer science
Haskell overloading is DEXPTIME-complete
Information Processing Letters
Co-definite set constraints with membership expressions
JICSLP'98 Proceedings of the 1998 joint international conference and symposium on Logic programming
Generalized Definite Set Constraints
Constraints
RTA '98 Proceedings of the 9th International Conference on Rewriting Techniques and Applications
LICS '98 Proceedings of the 13th Annual IEEE Symposium on Logic in Computer Science
| Hi-index | 0.00 |
The Horn µ-calculus is a formalism extending logic programs by specifying for each predicate symbol with which (greatest or least) fix-point semantics, its denotation has to be computed. When restricted to a particular class of logic programs called uniform, the Horn µ-calculus provides a syntactic extension for Rabin tree automata. However, it has been shown [1] that the denotation of the Horn µ-calculus restricted to a uniform program remains a regular set of trees and that moreover, the emptiness of the denotation of a predicate p is a DEXPTIME-complete problem (in the size of the program). In [3], these results have been extended to uniform programs that may contain both existential and universal quantifications on the variables occurring in the body of "clauses": considering this extension, the denotation of a program remains a regular set of trees, but the best known algorithm for testing the emptiness of the denotation of a predicate is doubly-exponential in the size of the program. In this paper, we consider uniform logic programs with both kinds of quantification in the body. But we add to the Horn µ-calculus a limitation on the way the fix-point semantics is specified for predicates. This restriction is close to the one defining the alternation-free fragment of the µ-calculus. Therefore, we name this fragment of the Horn µ-calculus the alternation-free fragment. We devise for it an algorithm which performs the emptiness test for the denotation of a predicate in single-exponential time in the size of the program. To obtain this result, we develop a constructive approach based on a new kind of tree automata running on finite and infinite trees, called monotonous tree automata. These automata are defined by means of a family of finite and complete lattices. The acceptance condition for monotonous tree automata is based on the ordering relations of the lattices.