On OBDDs for CNFs of Bounded Treewidth

On OBDDs for CNFs of Bounded Treewidth
复制标题

关于有界树宽 CNF 的 OBDD

DOI:
--
复制
发表时间:
2013
期刊:
International Conference on Principles of Knowledge Representation and Reasoning
影响因子:
--
通讯作者:
Igor Razgon
Igor Razgon
中科院分区:
--
文献类型:
--
作者:
Igor Razgon

文献摘要

被引文献

相似文献

在本文中,我们表明,一个CNF不能编译成一个有序的二叉决策图(OBDD)的固定参数的大小参数化的原始图树宽的CNF。因此,我们提供了一个参数化的分离OBDD和句子决策图(SDDs),这种固定参数的编译是可能的。事实上,我们证明,建议的下限也产生一个经典的(非参数化)分离OBDD和SDDs。我们还表明,最好的OBDDs现有的参数化上限,事实上持有关联图树宽参数化。
In this paper we show that a CNF cannot be compiled into an Ordered Binary Decision Diagram (OBDD) of fixed-parameter size parameterized by the primal graph treewidth of the CNF. Thus we provide a parameterized separation between OBDDs and Sentential Decision Diagrams (SDDs) for which such fixed-parameter compilation is possible. In fact, we demonstrate that the proposed lower bound also yields a classical (non-parameterized) separation of OBDDs and SDDs. We also show that the best existing parameterized upper bound for OBDDs in fact holds for incidence graph treewidth parameterization.