Formulas In Normal Form Research Articles

The Complexity of Problems for Quantified Constraints

In this paper we are interested in quantified propositional formulas in conjunctive normal form with “clauses” of arbitrary shapes. i.e., consisting of applying arbitrary relations to variables. We study the complexity of the evaluation problem, the model checking problem, the equivalence problem, and the counting problem for such formulas, both with and without a bound on the number of quantifier alternations. For each of these computational goals we get full complexity classifications: We determine the complexity of each of these problems depending on the set of relations allowed in the input formulas. Thus, on the one hand we exhibit syntactic restrictions of the original problems that are still computationally hard, and on the other hand we identify non-trivial subcases that admit efficient algorithms.

Narrow Proofs May Be Spacious:Separating Space and Width in Resolution

The width of a resolution proof is the maximal number of literals in any clause of the proof. The space of a proof is the maximal number of clauses kept in memory simultaneously if the proof is only allowed to infer new clauses from clauses currently in memory. Both of these measures have previously been studied and related to the resolution refutation size of unsatisfiable conjunctive normal form (CNF) formulas. Also, the minimum refutation space of a formula has been proven to be at least as large as the minimum refutation width, but it has been open whether space can be separated from width or the two measures coincide asymptotically. We prove that there is a family of k-CNF formulas for which the refutation width in resolution is constant but the refutation space is nonconstant, thus solving a problem mentioned in several previous papers.

SymChaff: exploiting symmetry in a structure-aware satisfiability solver

This article presents a new low-overhead framework for representing and utilizing problem symmetry in propositional satisfiability algorithms. While many previous approaches have focused on symmetry extraction as a key component, the novelty in the proposed strategy lies in using high level problem description to pass on symmetry information to the SAT solver in a simple and concise form, in addition to the usual conjunctive normal form formula. This information, comprising of the so-called symmetry sets and variable classes, captures variable semantics relevant to “complete multi-class symmetry” and is utilized dynamically to prune the search space. This allows us to address many limitations of alternative approaches like symmetry breaking predicates, implicit pseudo-Boolean representations, general group-theoretic methods, and zero-suppressed binary decision diagrams. We demonstrate the efficacy of our technique through a solver called SymChaff, which achieves exponential speedup over one of the best systematic SAT solvers on problems from both theory and practice, often by simply using natural tags or annotation in the problem specification.

A solver for QBFs in negation normal form

Various problems in artificial intelligence can be solved by translating them into a quantified boolean formula (QBF) and evaluating the resulting encoding. In this approach, a QBF solver is used as a black box in a rapid implementation of a more general reasoning system. Most of the current solvers for QBFs require formulas in prenex conjunctive normal form as input, which makes a further translation necessary, since the encodings are usually not in a specific normal form. This additional step increases the number of variables in the formula or disrupts the formula's structure. Moreover, the most important part of this transformation, prenexing, is not deterministic. In this paper, we focus on an alternative way to process QBFs without these drawbacks and describe a solver, $\ensuremath{\sf qpro}$ , which is able to handle arbitrary formulas. To this end, we extend algorithms for QBFs to the non-normal form case and compare $\ensuremath{\sf qpro}$ with the leading normal form provers on several problems from the area of artificial intelligence. We prove properties of the algorithms generalized to non-clausal form by using a novel approach based on a sequent-style formulation of the calculus.

A New Proposal Of Quasi-Solved Form For Equality Constraint Solving

Most well-known algorithms for equational solving are based on quantifier elimination. This technique iteratively eliminates the innermost block of existential/universal quantifiers from prenex formulas whose matrices are in some normal form (mostly DNF). Traditionally used notions of normal form satisfy that every constraint (in normal form) different from false is trivially satisfiable. Hence, they are called solved forms. However, the manipulation of such constraints require hard transformations, especially due to the use of the distributive and the explosion rules, which increase the number of constraints at intermediate stages of the solving process. On the contrary, quasi-solved forms allow for simpler transformations by means of a more compact representation of solutions, but their satisfiability test is not so trivial. Nevertheless, the total cost of checking satisfiability and manipulating constrains using quasi-solved forms is cheaper than using simpler solved forms. Therefore, they are suitable for improving the efficiency of constraint solving procedures. In this paper, we present a notion of quasi-solved form that provides a good trade-off between the cost of checking satisfiability and the effort required to manipulate constraints. In particular, our new quasi-solved form has been carefully designed for efficiently handling conjunction and negation, which are the main Boolean operations necessary to keep matrices of formulas in normal form.

DNA computing approach for automated test pattern generation for digital circuits

Testing of digital circuits is a compute intensive problem. This article deals with the problem of automated test pattern generation for large digital circuits. A new evolutionary approach based on DNA computing is presented, which exploits the computational power of DNA molecules to solve the problem. A Boolean formula in conjunctive normal form is extracted from the circuit under test and then the proposed algorithm based on DNA computing is used to find the solution satisfying that formula. Exploiting the massive parallelism and recombination properties of DNA molecules, a test vector is found in polynomial time i.e., O (nk). Its effectiveness in terms of Fault coverage, CPU time and Test vector generated is compared with some existing classical approaches like simulated annealing and genetic algorithms.

SAT graph-based representation: A new perspective

In this paper a new graph based representation of Boolean formulas in conjunctive normal form (CNF) is proposed. It extends the well-known graph representation of binary CNF formulas (2-SAT) to the general case. Every clause is represented as a set of (conditional) implications and encoded with different edges labeled with a set of literals, called context. This representation admits many interesting features. For example, a path from the node labeled with a literal ¬ x to the node labeled with a literal x gives us an original way to compute the condition under which the literal x is implied. Using this representation, we show that classical resolution can be reformulated as a transitive closure on the generated graph. Interestingly enough, using the SAT graph-based representation three original applications are then derived. The first one deals with the 2-SAT strong backdoor set computation problem, whereas in the second one the underlying representation is used to derive hard SAT instances with respect to the state-of-the-art satisfiability solvers. Finally, a new preprocessing technique of CNF formulas which extends the well-known hyper-resolution rule is proposed. Experimental results show interesting improvements on many classes of SAT instances taken from the last SAT competitions.

An Extension of Deficiency and Minimal Unsatisfiability of Quantified Boolean Formulas

The notion of minimal unsatisfiability and minimal falsity will be extended to non-clausal and non-prenex quantified Boolean formulas. For quantified Boolean formulas in negation normal form we generalize the notion of deficiency to the so-called cohesion, which is 1+ the difference between the number of occurrences of the conjunction symbol ∧ and the number of existential and free variables. Further, we show that all the complexity results with respect to minimal unsatisfiability or minimal falsity known for formulas with fixed deficiency can be adapted to formulas with fixed cohesion. For example, (1) the minimal unsatisfiability of propositional formulas with fixed cohesion is still solvable in polynomial time and, (2) the minimal falsity of quantified Boolean formulas with cohesion 1 is solvable in polynomial time.

Solving #SAT using vertex covers

We propose an exact algorithm for counting the models of propositional formulas in conjunctive normal form. Our algorithm is based on the detection of strong backdoor sets of bounded size; each instantiation of the variables of a strong backdoor set puts the given formula into a class of formulas for which models can be counted in polynomial time. For the backdoor set detection we utilize an efficient vertex cover algorithm applied to a certain “obstruction graph” that we associate with the given formula. This approach gives rise to a new hardness index for formulas, the clustering-width. Our algorithm runs in uniform polynomial time on formulas with bounded clustering-width. It is known that the number of models of formulas with bounded clique-width, bounded treewidth, or bounded branchwidth can be computed in polynomial time; these graph parameters are applied to formulas via certain (hyper)graphs associated with formulas. We show that clustering-width and the other parameters mentioned are incomparable: there are formulas with bounded clustering-width and arbitrarily large clique-width, treewidth, and branchwidth. Conversely, there are formulas with arbitrarily large clustering-width and bounded clique-width, treewidth, and branchwidth.

Lower Bounds for Lovász–Schrijver Systems and Beyond Follow from Multiparty Communication Complexity

We prove that an $\omega(\log^4 n)$ lower bound for the three-party number-on-the-forehead (NOF) communication complexity of the set-disjointness function implies an $n^{\omega(1)}$ size lower bound for treelike Lovasz-Schrijver systems that refute unsatisfiable formulas in conjunctive normal form (CNFs). More generally, we prove that an $n^{\Omega(1)}$ lower bound for the $(k+1)$-party NOF communication complexity of set disjointness implies a $2^{n^{\Omega(1)}}$ size lower bound for all treelike proof systems whose formulas are degree $k$ polynomial inequalities.

Global bifurcations and chaotic dynamics for a string-beam coupled system

The global bifurcations and chaotic dynamics of a string-beam coupled system subjected to parametric and external excitations are investigated in detail in this paper. The governing equations are firstly obtained to describe the nonlinear transverse vibrations of the string-beam coupled system. The Galerkin procedure is introduced to simplify the governing equations of motion to ordinary differential equations with two-degrees-of-freedom. Using the method of multiple scales, parametrically and externally excited system is transformed to the averaged equation. The case of 1:2 internal resonance between the modes of the beam and string, principal parametric resonance for the beam and primary resonance for the string is considered. Based on the averaged equation, the theory of normal form is utilized to find the explicit formulas of normal form associated with one double zero and a pair of pure imaginary eigenvalues. The global perturbation method is employed to analyze the global bifurcations and chaotic dynamics of the string-beam coupled system. The analysis of the global bifurcations indicates that there exist the homoclinic bifurcations and the Silnikov type single-pulse homoclinic orbit in the averaged equation of the string-beam coupled system. These results obtained here mean that the chaotic motions can occur in the string-beam coupled system. Numerical simulations also verify the analytical predications.

Clause/Term Resolution and Learning in the Evaluation of Quantified Boolean Formulas

Resolution is the rule of inference at the basis of most procedures for automated reasoning. In these procedures, the input formula is first translated into an equisatisfiable formula in conjunctive normal form (CNF) and then represented as a set of clauses. Deduction starts by inferring new clauses by resolution, and goes on until the empty clause is generated or satisfiability of the set of clauses is proven, e.g., because no new clauses can be generated. In this paper, we restrict our attention to the problem of evaluating Quantified Boolean Formulas (QBFs). In this setting, the above outlined deduction process is known to be sound and complete if given a formula in CNF and if a form of resolution, called ``Q-resolution'', is used. We introduce Q-resolution on terms, to be used for formulas in disjunctive normal form. We show that the computation performed by most of the available procedures for QBFs --based on the Davis-Logemann-Loveland procedure (DLL) for propositional satisfiability-- corresponds to a tree in which Q-resolution on terms and clauses alternate. This poses the theoretical bases for the introduction of learning, corresponding to recording Q-resolution formulas associated with the nodes of the tree. We discuss the problems related to the introduction of learning in DLL based procedures, and present solutions extending state-of-the-art proposals coming from the literature on propositional satisfiability. Finally, we show that our DLL based solver extended with learning, performs significantly better on benchmarks used in the 2003 QBF solvers comparative evaluation.

Efficient symmetry breaking for Boolean satisfiability

Identifying and breaking the symmetries of conjunctive normal form (CNF) formulae has been shown to lead to significant reductions in search times. Symmetries in the search space are broken by adding appropriate symmetry-breaking predicates (SBPs) to an SAT instance in CNF. The SBPs prune the search space by acting as a filter that confines the search to nonsymmetric regions of the space without affecting the satisfiability of the CNF formula. For symmetry breaking to be effective in practice, the computational overhead of generating and manipulating SBPs must be significantly less than the runtime savings they yield due to search space pruning. In this paper, we describe a more systematic and efficient construction of SBPs. In particular, we use the cycle structure of symmetry generators, which typically involve very few variables, to drastically reduce the size of SBPs. Furthermore, our new SBP construction grows linearly with the number of relevant variables as opposed to the previous quadratic constructions. Our empirical data suggest that these improvements reduce search runtimes by one to two orders of magnitude on a wide variety of benchmarks with symmetries.

On the computational power of Boolean decision lists

We study the computational power of decision lists over AND-functions versus $$ \hbox{threshold-}\oplus $$ circuits. AND-decision lists are a natural generalization of formulas in disjunctive or conjunctive normal form. We show that, in contrast to CNF- and DNF-formulas, there are functions with small AND-decision lists which need exponential size unbounded weight $$\hbox{threshold-}\oplus $$ circuits. Consequently, it is questionable if the polynomial learning algorithm for DNFs of Jackson (1994), which is based on the efficient simulation of DNFs by polynomial weight $$ \hbox{threshold-}\oplus $$ circuits (Krause & Pudlak 1994), can be successfully applied to functions with small AND-decision lists. A further result is that for all k ? 1 the complexity class defined by polynomial length AC0k-decision lists lies strictly between AC0k+1 and AC0k+2.

Conformant planning via heuristic forward search: A new approach

Conformant planning is the task of generating plans given uncertainty about the initial state and action effects, and without any sensing capabilities during plan execution. The plan should be successful regardless of which particular initial world we start from. It is well known that conformant planning can be transformed into a search problem in belief space, the space whose elements are sets of possible worlds. We introduce a new representation of that search space, replacing the need to store sets of possible worlds with a need to reason about the effects of action sequences. The reasoning is done by implication tests on propositional formulas in conjunctive normal form (CNF) that capture the action sequence semantics. Based on this approach, we extend the classical heuristic forward-search planning system FF to the conformant setting. The key to this extension is an appropriate extension of the relaxation that underlies FF's heuristic function, and of FF's machinery for solving relaxed planning problems: the extended machinery includes a stronger form of the CNF implication tests that we use to reason about the effects of action sequences. Our experimental evaluation shows the resulting planning system to be superior to the state-of-the-art conformant planners MBP, KACMBP, and GPT in a variety of benchmark domains.

Applications of Minimal Unsatisfiable Formulas to Polynomially Reduction for Formulas

A conjunctive normal form (CNF) formula F is minimal unsatisfiable if F is unsatisfiable and the resulting formula removing any clause from F is satisfiable. (r,s)-CNF is a subclass of CNF in which each clause of formula has exactly r distinct literals and every variable occurs at most s times. The corresponding satisfiable problem (r,s)-SAT means that the instances are restricted in (r,s)-CNF. For positive integer r≥3, there exists a critical function f(r) such that all formulas in (r,f(r))-CNF are satisfiable, but (r,f(r)+1)-SAT is already NP-Complete. It is open whether or not the function f is computable. One can only estimate some bounds of f(r) except for f(3)=3 and f(4)=4. In this paper, the applications of minimal unsatisfiable formulas are described for transformations between CNF formulas. A new algorithm is presented to introduce less new variables in transformation from CNF to 3-CNF, which for clauses with length l(>3) only     2 l new variables are introduced. The principle and method for transforming CNF to (r,s)-CNF and constructing unsatisfiable formulas in (r,s)-CNF are presented.

A Faster Clause-Shortening Algorithm for SAT with No Restriction on Clause Length

We give a randomized algorithm for testing satisfiability of Boolean formulas in conjunctive normal form with no restriction on clause length. This algorithm uses the clause-shortening approach proposed by Schuler. The running time of the algorithm is O(2^(n(1−1/α))) where α = ln(m/n) + O(ln ln m) and n, m are respectively the number of variables and the number of clauses in the input formula. This bound is asymptotically better than the previously best known 2^(n(1−1/ log(2m))) bound for SAT.

The complexity of maximum matroid–greedoid intersection and weighted greedoid maximization

The maximum intersection problem for a matroid and a greedoid, given by polynomial-time oracles, is shown NP-hard by expressing the satisfiability of boolean formulas in 3-conjunctive normal form as such an intersection. The corresponding approximation problems are shown NP-hard for certain approximation performance bounds. Moreover, some natural parameterized variants of the problem are shown W[P]-hard. The results are in contrast with the maximum matroid–matroid intersection which is solvable in polynomial time by an old result of Edmonds. We also prove that it is NP-hard to approximate the weighted greedoid maximization within 2 n O ( 1 ) where n is the size of the domain of the greedoid.

Efficiency versus Convergence of Boolean Kernels for On-Line Learning Algorithms

The paper studies machine learning problems where each example is described using a set of Boolean features and where hypotheses are represented by linear threshold elements. One method of increasing the expressiveness of learned hypotheses in this context is to expand the feature set to include conjunctions of basic features. This can be done explicitly or where possible by using a kernel function. Focusing on the well known Perceptron and Winnow algorithms, the paper demonstrates a tradeoff between the computational efficiency with which the algorithm can be run over the expanded feature space and the generalization ability of the corresponding learning algorithm. We first describe several kernel functions which capture either limited forms of conjunctions or all conjunctions. We show that these kernels can be used to efficiently run the Perceptron algorithm over a feature space of exponentially many conjunctions; however we also show that using such kernels, the Perceptron algorithm can provably make an exponential number of mistakes even when learning simple functions. We then consider the question of whether kernel functions can analogously be used to run the multiplicative-update Winnow algorithm over an expanded feature space of exponentially many conjunctions. Known upper bounds imply that the Winnow algorithm can learn Disjunctive Normal Form (DNF) formulae with a polynomial mistake bound in this setting. However, we prove that it is computationally hard to simulate Winnow's behavior for learning DNF over such a feature set. This implies that the kernel functions which correspond to running Winnow for this problem are not efficiently computable, and that there is no general construction that can run Winnow with kernels.

CASCADING RANDOM WALKS

In this paper, we discuss a simple, Monte Carlo algorithm for the problem of checking whether a Quantified Boolean Formula (QBF) in Conjunctive Normal Form (CNF), with at most two literals per clause has a model. The term k-CNF is used to describe boolean formulas in CNF, with at most k literals per clause and the problem of checking whether a given k-CNF formula is satisfiable is called the k-SAT problem. A QBF is a boolean formula, accompanied by a quantifier string which imposes a linear ordering on the variables of that formula. The problem of finding a model for a QBF formula in CNF, with at at most k literals per clause is called the QkSAT problem. The QkSAT problem is PSPACE-complete, for k≥3. However, the Q2SAT problem can be decided in polynomial time; the graph-based procedure, discussed in [1], is the first such algorithm for this problem. This procedure requires the construction of a global implication graph, corresponding to the input formula and searching for certain paths in this graph. Hence the complete set of clauses must be part of the input. We propose an incremental, randomized approach for the Q2SAT problem that is essentially local in nature, in that the complete clausal set need not be provided at any time, in the presence of a verifier. We show that the randomized algorithm can be analyzed as a one-dimensional random walk, with one reflecting barrier and one absorbing barrier. On a Q2SAT instance with m clauses on n variables, our coin-flipping algorithm runs in time O(n2 · V(m, n)), where V(m, n) is the time required to verify that a given model satisfies the formula. Additionally, if the instance is satisfiable, the probability that our algorithm fails to find a model is less than one half. The design and analysis of a randomized algorithm for a problem, is important from both the theoretical and the practical perspectives. Randomized approaches tend to be simple and elegant, thereby making the process of checking correctness, effortless as well. The randomized approach discussed in this paper lays the groundwork for analyzing a number of problems related to 2CNF formulas and directed graphs. We remark that our work in this paper is the first randomized algorithm for a class of QBFs.