Bounded second-order unification is NP-complete

  • Authors:
  • Jordi Levy;Manfred Schmidt-Schauß;Mateu Villaret

  • Affiliations:
  • IIIA, CSIC, Barcelona, Spain;Institut für Informatik, Johann Wolfgang Goethe-Universität, Frankfurt, Germany;IMA, UdG, Girona, Spain

  • Venue:
  • RTA'06 Proceedings of the 17th international conference on Term Rewriting and Applications
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Bounded Second-Order Unification is the problem of deciding, for a given second-order equation ${t {\stackrel{_?}=} u}$ and a positive integer m, whether there exists a unifier σ such that, for every second-order variable F, the terms instantiated for F have at most m occurrences of every bound variable. It is already known that Bounded Second-Order Unification is decidable and NP-hard, whereas general Second-Order Unification is undecidable. We prove that Bounded Second-Order Unification is NP-complete, provided that m is given in unary encoding, by proving that a size-minimal solution can be represented in polynomial space, and then applying a generalization of Plandowski's polynomial algorithm that compares compacted terms in polynomial time.