arXiv Analytics

Sign in

arXiv:1807.04263 [cs.CC]AbstractReferencesReviewsResources

Knowledge Compilation, Width and Quantification

Florent Capelli, Stefan Mengel

Published 2018-07-11Version 1

We generalize many results concerning the tractability of SAT and #SAT on bounded treewidth CNF-formula in the context of Quantified Boolean Formulas (QBF). To this end, we start by studying the notion of width for OBDD and observe that the blow up in size while existentially or universally projecting a block of variables in an OBDD only affects its width. We then generalize this notion of width to the more general representation of structured (deterministic) DNNF and give a similar algorithm to existentially or universally project a block of variables. Using a well-known algorithm transforming bounded treewidth CNF formula into deterministic DNNF, we are able to generalize this connection to quantified CNF which gives us as a byproduct that one can count the number of models of a bounded treewidth and bounded quantifier alternation quantified CNF in FPT time. We also give an extensive study of bounded width d-DNNF and proves the optimality of several of our results.

Related articles:
arXiv:1701.01461 [cs.CC] (Published 2017-01-05)
Understanding the complexity of #SAT using knowledge compilation