Arrow Research search
Back to AAAI

AAAI 2000

Tractable Classes for Directional Resolution

Conference Paper Computational Complexity of Reasoning Artificial Intelligence

Abstract

The original, resolution-based Davis-Putnam satisfiability algorithm (Davis & Putnam 1960) was recently revived by (Dechter & Rish 1994) under the name “directional resolution” (DR). We provide new positive complexity results for DR. First, we identify a class of theories (ACT, Acyclic Component Theories), which includes many real-world theories, for which DR takes polynomial time. Second, we present an improved analysis of the complexity of directional resolution through refined notions of induced width, which yields new tractable classes for DR, and much better predictions of its space and time requirements under various atom orderings. These estimates can be used for heuristically choosing among various orderings before running DR.

Authors

Keywords

No keywords are indexed for this paper.

Context

Venue
AAAI Conference on Artificial Intelligence
Archive span
1980-2026
Indexed papers
28718
Paper id
1097268401745918464