arXiv subjects

Miroslaw Truszczynski

Publications and source records attributed to Miroslaw Truszczynski.

At least 37 records · Page 2Linked to original sources

arXiv2003

Local-search techniques for propositional logic extended with cardinality constraints

We study local-search satisfiability solvers for propositional logic extended with cardinality atoms, that is, expressions that provide explicit ways to model constraints on cardinalities of sets. Adding cardinality atoms to the language of propositional logic facilitates modeling search problems and often results in concise encodings. We propose two ``native'' local-search solvers for theories in the extended language. We also describe techniques to reduce the problem to standard propositional satisfiability and allow us to use off-the-shelf SAT solvers. We study these methods experimentally. Our general finding is that native solvers designed specifically for the extended language perform better than indirect methods relying on SAT solvers.

cs.AI↗

arXiv2003

WSAT(cc) - a fast local-search ASP solver

We describe WSAT(cc), a local-search solver for computing models of theories in the language of propositional logic extended by cardinality atoms. WSAT(cc) is a processing back-end for the logic PS+, a recently proposed formalism for answer-set programming.

cs.AI↗

arXiv2003

Logic programs with monotone cardinality atoms

We investigate mca-programs, that is, logic programs with clauses built of monotone cardinality atoms of the form kX, where k is a non-negative integer and X is a finite set of propositional atoms. We develop a theory of mca-programs. We demonstrate that the operational concept of the one-step provability operator generalizes to mca-programs, but the generalization involves nondeterminism. Our main results show that the formalism of mca-programs is a common generalization of (1) normal logic programming with its semantics of models, supported models and stable models, (2) logic programming with cardinality atoms and with the semantics of stable models, as defined by Niemela, Simons and Soininen, and (3) of disjunctive logic programming with the possible-model semantics of Sakama and Inoue.

cs.LO↗

arXiv2003

Satisfiability and computing van der Waerden numbers

In this paper we bring together the areas of combinatorics and propositional satisfiability. Many combinatorial theorems establish, often constructively, the existence of positive integer functions, without actually providing their closed algebraic form or tight lower and upper bounds. The area of Ramsey theory is especially rich in such results. Using the problem of computing van der Waerden numbers as an example, we show that these problems can be represented by parameterized propositional theories in such a way that decisions concerning their satisfiability determine the numbers (function) in question. We show that by using general-purpose complete and local-search techniques for testing propositional satisfiability, this approach becomes effective -- competitive with specialized approaches. By following it, we were able to obtain several new results pertaining to the problem of computing van der Waerden numbers. We also note that due to their properties, especially their structural simplicity and computational hardness, propositional theories that arise in this research can be of use in development, testing and benchmarking of SAT solvers.

cs.LO↗

arXiv2002

Propositional satisfiability in declarative programming

Answer-set programming (ASP) paradigm is a way of using logic to solve search problems. Given a search problem, to solve it one designs a theory in the logic so that models of this theory represent problem solutions. To compute a solution to a problem one needs to compute a model of the corresponding theory. Several answer-set programming formalisms have been developed on the basis of logic programming with the semantics of stable models. In this paper we show that also the logic of predicate calculus gives rise to effective implementations of the ASP paradigm, similar in spirit to logic programming with stable model semantics and with a similar scope of applicability. Specifically, we propose two logics based on predicate calculus as formalisms for encoding search problems. We show that the expressive power of these logics is given by the class NP-search. We demonstrate how to use them in programming and develop computational tools for model finding. In the case of one of the logics our techniques reduce the problem to that of propositional satisfiability and allow one to use off-the-shelf satisfiability solvers. The language of the other logic has more complex syntax and provides explicit means to model some high-level constraints. For theories in this logic, we designed our own solver that takes advantage of the expanded syntax. We present experimental results demonstrating computational effectiveness of the overall approach.

cs.LO↗

arXiv2002

Computing stable models: worst-case performance estimates

We study algorithms for computing stable models of propositional logic programs and derive estimates on their worst-case performance that are asymptotically better than the trivial bound of O(m 2^n), where m is the size of an input program and n is the number of its atoms. For instance, for programs, whose clauses consist of at most two literals (counting the head) we design an algorithm to compute stable models that works in time O(m\times 1.44225^n). We present similar results for several broader classes of programs, as well.

cs.LO↗

arXiv2002

Ultimate approximations in nonmonotonic knowledge representation systems

We study fixpoints of operators on lattices. To this end we introduce the notion of an approximation of an operator. We order approximations by means of a precision ordering. We show that each lattice operator O has a unique most precise or ultimate approximation. We demonstrate that fixpoints of this ultimate approximation provide useful insights into fixpoints of the operator O. We apply our theory to logic programming and introduce the ultimate Kripke-Kleene, well-founded and stable semantics. We show that the ultimate Kripke-Kleene and well-founded semantics are more precise then their standard counterparts We argue that ultimate semantics for logic programming have attractive epistemological properties and that, while in general they are computationally more complex than the standard semantics, for many classes of theories, their complexity is no worse.

cs.AI↗

arXiv2001

Fixed-parameter complexity of semantics for logic programs

A decision problem is called parameterized if its input is a pair of strings. One of these strings is referred to as a parameter. The problem: given a propositional logic program P and a non-negative integer k, decide whether P has a stable model of size no more than k, is an example of a parameterized decision problem with k serving as a parameter. Parameterized problems that are NP-complete often become solvable in polynomial time if the parameter is fixed. The problem to decide whether a program P has a stable model of size no more than k, where k is fixed and not a part of input, can be solved in time O(mn^k), where m is the size of P and n is the number of atoms in P. Thus, this problem is in the class P. However, algorithms with the running time given by a polynomial of order k are not satisfactory even for relatively small values of k. The key question then is whether significantly better algorithms (with the degree of the polynomial not dependent on k) exist. To tackle it, we use the framework of fixed-parameter complexity. We establish the fixed-parameter complexity for several parameterized decision problems involving models, supported models and stable models of logic programs. We also establish the fixed-parameter complexity for variants of these problems resulting from restricting attention to Horn programs and to purely negative programs. Most of the problems considered in the paper have high fixed-parameter complexity. Thus, it is unlikely that fixing bounds on models (supported models, stable models) will lead to fast algorithms to decide the existence of such models.

cs.LO↗

arXiv2001

Annotated revision programs

Revision programming is a formalism to describe and enforce updates of belief sets and databases. That formalism was extended by Fitting who assigned annotations to revision atoms. Annotations provide a way to quantify the confidence (probability) that a revision atom holds. The main goal of our paper is to reexamine the work of Fitting, argue that his semantics does not always provide results consistent with intuition, and to propose an alternative treatment of annotated revision programs. Our approach differs from that proposed by Fitting in two key aspects: we change the notion of a model of a program and we change the notion of a justified revision. We show that under this new approach fundamental properties of justified revisions of standard revision programs extend to the annotated case.

cs.AI↗

arXiv2001

Propositional satisfiability in answer-set programming

We show that propositional logic and its extensions can support answer-set programming in the same way stable logic programming and disjunctive logic programming do. To this end, we introduce a logic based on the logic of propositional schemata and on a version of the Closed World Assumption. We call it the extended logic of propositional schemata with CWA (PS+, in symbols). An important feature of this logic is that it supports explicit modeling of constraints on cardinalities of sets. In the paper, we characterize the class of problems that can be solved by finite PS+ theories. We implement a programming system based on the logic PS+ and design and implement a solver for processing theories in PS+. We present encouraging performance results for our approach --- we show it to be competitive with smodels, a state-of-the-art answer-set programming system based on stable logic programming.

cs.AI↗

arXiv2001

On the problem of computing the well-founded semantics

The well-founded semantics is one of the most widely studied and used semantics of logic programs with negation. In the case of finite propositional programs, it can be computed in polynomial time, more specifically, in O(|At(P)|size(P)) steps, where size(P) denotes the total number of occurrences of atoms in a logic program P. This bound is achieved by an algorithm introduced by Van Gelder and known as the alternating-fixpoint algorithm. Improving on the alternating-fixpoint algorithm turned out to be difficult. In this paper we study extensions and modifications of the alternating-fixpoint approach. We then restrict our attention to the class of programs whose rules have no more than one positive occurrence of an atom in their bodies. For programs in that class we propose a new implementation of the alternating-fixpoint method in which false atoms are computed in a top-down fashion. We show that our algorithm is faster than other known algorithms and that for a wide class of programs it is linear and so, asymptotically optimal.

cs.LO↗

arXiv2000

Computing large and small stable models

In this paper, we focus on the problem of existence and computing of small and large stable models. We show that for every fixed integer k, there is a linear-time algorithm to decide the problem LSM (large stable models problem): does a logic program P have a stable model of size at least |P|-k. In contrast, we show that the problem SSM (small stable models problem) to decide whether a logic program P has a stable model of size at most k is much harder. We present two algorithms for this problem but their running time is given by polynomials of order depending on k. We show that the problem SSM is fixed-parameter intractable by demonstrating that it is W[2]-hard. This result implies that it is unlikely, an algorithm exists to compute stable models of size at most k that would run in time O(n^c), where c is a constant independent of k. We also provide an upper bound on the fixed-parameter complexity of the problem SSM by showing that it belongs to the class W[3].

cs.LO↗

arXiv2000

DATALOG with constraints - an answer-set programming system

Answer-set programming (ASP) has emerged recently as a viable programming paradigm well attuned to search problems in AI, constraint satisfaction and combinatorics. Propositional logic is, arguably, the simplest ASP system with an intuitive semantics supporting direct modeling of problem constraints. However, for some applications, especially those requiring that transitive closure be computed, it requires additional variables and results in large theories. Consequently, it may not be a practical computational tool for such problems. On the other hand, ASP systems based on nonmonotonic logics, such as stable logic programming, can handle transitive closure computation efficiently and, in general, yield very concise theories as problem representations. Their semantics is, however, more complex. Searching for the middle ground, in this paper we introduce a new nonmonotonic logic, DATALOG with constraints or DC. Informally, DC theories consist of propositional clauses (constraints) and of Horn rules. The semantics is a simple and natural extension of the semantics of the propositional logic. However, thanks to the presence of Horn rules in the system, modeling of transitive closure becomes straightforward. We describe the syntax and semantics of DC, and study its properties. We discuss an implementation of DC and present results of experimental study of the effectiveness of DC, comparing it with CSAT, a satisfiability checker and SMODELS implementation of stable logic programming. Our results show that DC is competitive with the other two approaches, in case of many search problems, often yielding much more efficient solutions.

cs.AI↗

arXiv2000

Proceedings of the 8th International Workshop on Non-Monotonic Reasoning, NMR'2000

The papers gathered in this collection were presented at the 8th International Workshop on Nonmonotonic Reasoning, NMR2000. The series was started by John McCarthy in 1978. The first international NMR workshop was held at Mohonk Mountain House, New Paltz, New York in June, 1984, and was organized by Ray Reiter and Bonnie Webber. In the last 10 years the area of nonmonotonic reasoning has seen a number of important developments. Significant theoretical advances were made in the understanding of general abstract principles underlying nonmonotonicity. Key results on the expressibility and computational complexity of nonmonotonic logics were established. The role of nonmonotonic reasoning in belief revision, abduction, reasoning about action, planing and uncertainty was further clarified. Several successful NMR systems were built and used in applications such as planning, scheduling, logic programming and constraint satisfaction. The papers in the proceedings reflect these recent advances in the field. They are grouped into sections corresponding to special sessions as they were held at the workshop: 1. General NMR track 2. Abductive reasonig 3. Belief revision: theory and practice 4. Representing action and planning 5. Systems descriptions and demonstrations 6. Uncertainty frameworks in NMR

cs.AI↗

arXiv2000

dcs: An Implementation of DATALOG with Constraints

Answer-set programming (ASP) has emerged recently as a viable programming paradigm. We describe here an ASP system, DATALOG with constraints or DC, based on non-monotonic logic. Informally, DC theories consist of propositional clauses (constraints) and of Horn rules. The semantics is a simple and natural extension of the semantics of the propositional logic. However, thanks to the presence of Horn rules in the system, modeling of transitive closure becomes straightforward. We describe the syntax, use and implementation of DC and provide experimental results.

cs.AI↗

arXiv2000

On the accuracy and running time of GSAT

Randomized algorithms for deciding satisfiability were shown to be effective in solving problems with thousands of variables. However, these algorithms are not complete. That is, they provide no guarantee that a satisfying assignment, if one exists, will be found. Thus, when studying randomized algorithms, there are two important characteristics that need to be considered: the running time and, even more importantly, the accuracy --- a measure of likelihood that a satisfying assignment will be found, provided one exists. In fact, we argue that without a reference to the accuracy, the notion of the running time for randomized algorithms is not well-defined. In this paper, we introduce a formal notion of accuracy. We use it to define a concept of the running time. We use both notions to study the random walk strategy GSAT algorithm. We investigate the dependence of accuracy on properties of input formulas such as clause-to-variable ratio and the number of satisfying assignments. We demonstrate that the running time of GSAT grows exponentially in the number of variables of the input formula for randomly generated 3-CNF formulas and for the formulas encoding 3- and 4-colorability of graphs.

cs.AI↗

arXiv2000

Uniform semantic treatment of default and autoepistemic logics

We revisit the issue of connections between two leading formalisms in nonmonotonic reasoning: autoepistemic logic and default logic. For each logic we develop a comprehensive semantic framework based on the notion of a belief pair. The set of all belief pairs together with the so called knowledge ordering forms a complete lattice. For each logic, we introduce several semantics by means of fixpoints of operators on the lattice of belief pairs. Our results elucidate an underlying isomorphism of the respective semantic constructions. In particular, we show that the interpretation of defaults as modal formulas proposed by Konolige allows us to represent all semantics for default logic in terms of the corresponding semantics for autoepistemic logic. Thus, our results conclusively establish that default logic can indeed be viewed as a fragment of autoepistemic logic. However, as we also demonstrate, the semantics of Moore and Reiter are given by different operators and occupy different locations in their corresponding families of semantics. This result explains the source of the longstanding difficulty to formally relate these two semantics. In the paper, we also discuss approximating skeptical reasoning with autoepistemic and default logics and establish constructive principles behind such approximations.

cs.AI↗

arXiv1999

Representation Theory for Default Logic

Default logic can be regarded as a mechanism to represent families of belief sets of a reasoning agent. As such, it is inherently second-order. In this paper, we study the problem of representability of a family of theories as the set of extensions of a default theory. We give a complete solution to the representability by means of normal default theories. We obtain partial results on representability by arbitrary default theories. We construct examples of denumerable families of non-including theories that are not representable. We also study the concept of equivalence between default theories.

cs.LO↗