[Zurück]

F. Lonsing, U. Egly:
"Evaluating QBF Solvers: Quantifier Alternations Matter";
Vortrag: 24th International Conference on Principles and Practice of Constraint Programming (CP 2018), Lille, France; 27.08.2018 - 31.08.2018; in: "Principles and Practice of Constraint Programming - 24th International Conference, CP 2018, Lille, France, August 27-31, 2018, Proceedings", Springer, 11008 (2018), S. 276 - 294.

We present an experimental study of the effects of quantifier alternations on the evaluation of quantified Boolean formula (QBF) solvers. The number of quantifier alternations in a QBF in prenex conjunctive normal form (PCNF) is directly related to the theoretical hardness of the respective QBF satisfiability problem in the polynomial hierarchy. We show empirically that the performance of solvers based on different solving paradigms substantially varies depending on the numbers of alternations in PCNFs. In related theoretical work, quantifier alternations have become the focus of understanding the strengths and weaknesses of various QBF proof systems implemented in solvers. Our results motivate the development of methods to evaluate orthogonal solving paradigms by taking quantifier alternations into account. This is necessary to showcase the broad range of existing QBF solving paradigms for practical QBF applications. Moreover, we highlight the potential of combining different approaches and QBF proof systems in solvers.

http://dx.doi.org/10.1007/978-3-319-98334-9_19

https://link.springer.com/content/pdf/10.1007%2F978-3-319-98334-9_19.pdf