DOI

The partial string avoidability problem, also known as partial word avoidability, is stated as follows: Given a finite set of strings with possible "holes" (undefined symbols), determine whether there exists any two-sided infinite string containing no substrings from this set, assuming that a hole matches every symbol. The problem is known to be NP-hard and in PSPACE, and this paper establishes its PSPACE-completeness. Next, string avoidability over the binary alphabet is interpreted as a version of conjunctive normal form (CNF) satisfiability problem (SAT), with each clause having infinitely many shifted variants. Non-satisfiability of these formulas can be proved using variants of classical propositional proof systems, augmented with derivation rules for shifting constraints (such as clauses, inequalities, polynomials, etc). Two results on their proof complexity are established. First, there is a particular formula that has a short refutation in Resolution with shift, but requires classical proofs of exponential size (Resolution, Cutting Plane, Polynomial Calculus, etc.). At the same time, exponential lower bounds for shifted versions of classical proof systems are established.

Язык оригиналаанглийский
Название основной публикации41st International Symposium on Mathematical Foundations of Computer Science, MFCS 2016
РедакторыAnca Muscholl, Piotr Faliszewski, Rolf Niedermeier
ИздательSchloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing
ISBN (электронное издание)9783959770163
DOI
СостояниеОпубликовано - 1 авг 2016
Событие41st International Symposium on Mathematical Foundations of Computer Science, MFCS 2016 - Krakow, Польша
Продолжительность: 22 авг 201626 авг 2016

Серия публикаций

НазваниеLeibniz International Proceedings in Informatics, LIPIcs
Том58
ISSN (печатное издание)1868-8969

конференция

конференция41st International Symposium on Mathematical Foundations of Computer Science, MFCS 2016
Страна/TерриторияПольша
ГородKrakow
Период22/08/1626/08/16

    Предметные области Scopus

  • Программный продукт

ID: 41137701