Generalized Satisfiability with Limited Occurrences per Variable: A Study through Delta-Matroid Parity
In this paper we examine generalized satisfiability problems with limited variable occurrences. First, we show that 3 occurrences per variable suffice to make these problems as hard as their unrestricted version. Then we focus on generalized satisfiability problems with at most 2 occurrences per variable. It is known that some NP -complete generalized satisfiability problems become polynomially solvable when only 2 occurrences per variable are allowed. We identify two new families of generalized satisfiability problems, called local and binary, that are polynomially solvable when only 2 occurrences per variable are allowed. We achieve this result by means of a reduction to the \(\triangle\)-matroid parity problem, which is another important theme of this work.
KeywordsPolynomial Time Atomic Formula Graph Match Local Path Variable Occurrence
Unable to display preview. Download preview PDF.
- 3.Bouchet, A.: Representability of Δ-matroids. Combinatorics, pp. 167–182 (1988)Google Scholar
- 4.Bouchet, A.: Coverings and delta-coverings. Integer programming and combinatorial optimization, Copenhagen, pp. 228–243 (1995)Google Scholar
- 8.Garey, M.R., Johnson, D.S.: Computers and intractability (1979) Google Scholar
- 9.Geelen, J., Iwata, S., Murota, K.: The linear delta-matroid parity problem. Technical Report RIMS Preprint 1149, Kyoto University (1997) Google Scholar
- 10.Istrate, G.I.: Looking for a version of schaeer’s dichotomy theorem when each variable occurs at most twice. Technical Report TR652, The University of Rochester (March 1997) Google Scholar
- 13.Lovász, L., Plummer, M.D.: Matching Theory (1986) Google Scholar
- 15.Lovász, L.: The matroid matching problem. Algebraic methods in graph theory, Szeged, vol. I, II, pp. 495–517 (1978)Google Scholar
- 16.Schaefer, T.J.: The complexity of satisfiability problems. Theory of Computing (1978)Google Scholar