Limitations of Skeptical Default Reasoning

Jens Doerpmund

DOI: https://doi.org/10.48550/arXiv.1302.1536

IF: 14.4

2013-02-06

Artificial Intelligence

Abstract:Poole has shown that nonmonotonic logics do not handle the lottery paradox correctly. In this paper we will show that Pollock's theory of defeasible reasoning fails for the same reason: defeasible reasoning is incompatible with the skeptical notion of derivability.

What problem does this paper attempt to address?

A logic for default reasoning

R. Reiter

DOI: https://doi.org/10.1016/0004-3702(80)90014-4

IF: 14.4

1980-04-01

Artificial Intelligence

Abstract:The need to make default assumptions is frequently encountered in reasoning about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non-monotonicity of any logic of defaults.In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occuring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

computer science, artificial intelligence
Axiomatic Aspects of Default Inference

Guo-Qiang Zhang

DOI: https://doi.org/10.48550/arXiv.cs/0207087

2002-07-26

Abstract:This paper studies axioms for nonmonotonic consequences from a semantics-based point of view, focusing on a class of mathematical structures for reasoning about partial information without a predefined syntax/logic. This structure is called a default structure. We study axioms for the nonmonotonic consequence relation derived from extensions as in Reiter's default logic, using skeptical reasoning, but extensions are now used for the construction of possible worlds in a default information structure. In previous work we showed that skeptical reasoning arising from default-extensions obeys a well-behaved set of axioms including the axiom of cautious cut. We show here that, remarkably, the converse is also true: any consequence relation obeying this set of axioms can be represented as one constructed from skeptical reasoning. We provide representation theorems to relate axioms for nonmonotonic consequence relation and properties about extensions, and provide one-to-one correspondence between nonmonotonic systems which satisfies the law of cautious monotony and default structures with unique extensions. Our results give a theoretical justification for a set of basic rules governing the update of nonmonotonic knowledge bases, demonstrating the derivation of them from the more concrete and primitive construction of extensions. It is also striking to note that proofs of the representation theorems show that only shallow extensions are necessary, in the sense that the number of iterations needed to achieve an extension is at most three. All of these developments are made possible by taking a more liberal view of consistency: consistency is a user defined predicate, satisfying some basic properties.

Logic in Computer Science
Defeasible deontic logic: arguing about permission and obligation

Huimin Dong,Beishui Liao,Reka Markovich,Leendert van der Torre

2022-01-01

Abstract:Defeasible deontic logic uses techniques from non-monotonic logic to address various challenges in normative reasoning, such as prima facie permissions and obligations, moral dilemmas, deontic detachment, contrary-to-duty reasoning and legal interpretation. In this article, we use formal argumentation to design defeasible deontic logics, based on two classical deontic logics. In particular, we use the ASPIC(+) structured argumentation theory to define non-monotonic variants of well-understood monotonic modal logics. We illustrate the ASPIC(+)based approach and the resulting defeasible deontic logics using argumentation about strong permission.
Going Forth and Drawing Back: an Intensional Approach in Nonmonotonic Inference.

Yi Mao,Beihai Zhou,Beishui Liao

2015-01-01

Abstract:We decompose a nonmonotonic inference Gamma (sic) alpha into two stages: going forth to deduce all default conclusions in a logic system named DC (short for Default Conclusions), and drawing back less preferable conclusions in the face of conflicting default conclusions, based on a binary "more preferable" relation defined on those subformulas of that are also its deductive consequences. Under the possible-world semantics framework, we construct a set selection function and build up a theory of semantics, as a variant of traditional selection function semantics. We prove that the underlying logic system DC is sound and complete with respect to set selection function semantics. Using this two-layer mechanism, we account for benchmark examples including the Nixon Diamond and the Penguin Principle.
Decision under Insufficient Evidence: A Scalable Probabilistic Way

Xiaoqing Zheng,Hongjun Zhang,Feng Zhou

DOI: https://doi.org/10.1109/skg.2010.71

2010-01-01

Abstract:Some problematic cases, such as collective defeat and odd-length defeat cycles, which tend to be handled incorrectly by all of the current theories of no monotonic reasoning, including default logic and circumscription, have been well recognized in the literature. Although a powerful argument-based approach in the automated defeasible reasoner OSCAR has been proposed and they claim that this theory is able to reason correctly for the problems above all, but we don't consider it to be true completely through careful investigation. It seems to be the consequences of disconnection between epistemic reasoning and practical reasoning and not considering the possible consequences of the decision and individual preferences sufficiently. Following these observations, we propose a scalable probabilistic approach based on Bayesian decision theory that can solve all of the above paradoxes properly and has successfully been used in web of trust and knowledge integration for semantic Grid.
Reasoning about Exceptions in Ontologies: from the Lexicographic Closure to the Skeptical Closure

Laura Giordano,Valentina Gliozzi

DOI: https://doi.org/10.3233/fi-2020-1973

2020-12-18

Fundamenta Informaticae

Abstract:Reasoning about exceptions in ontologies is nowadays one of the challenges the description logics community is facing. The paper describes a preferential approach for dealing with exceptions in Description Logics, based on the rational closure. The rational closure has the merit of providing a simple and efficient approach for reasoning with exceptions, but it does not allow independent handling of the inheritance of different defeasible properties of concepts. In this work we outline a possible solution to this problem by introducing a weaker variant of the lexicographical closure, that we call skeptical closure, which requires to construct a single base. We develop a bi-preference semantics for defining a characterization of the skeptical closure.

computer science, software engineering,mathematics, applied
Non-Monotonicity in Probabilistic Reasoning

Benjamin N. Grosof

DOI: https://doi.org/10.48550/arXiv.1304.3087

2013-03-28

Abstract:We start by defining an approach to non-monotonic probabilistic reasoning in terms of non-monotonic categorical (true-false) reasoning. We identify a type of non-monotonic probabilistic reasoning, akin to default inheritance, that is commonly found in practice, especially in "evidential" and "Bayesian" reasoning. We formulate this in terms of the Maximization of Conditional Independence (MCI), and identify a variety of applications for this sort of default. We propose a formalization using Pointwise Circumscription. We compare MCI to Maximum Entropy, another kind of non-monotonic principle, and conclude by raising a number of open questions

Artificial Intelligence
Uniform semantic treatment of default and autoepistemic logics

Marc Denecker,Victor W. Marek,Miroslaw Truszczynski

DOI: https://doi.org/10.48550/arXiv.cs/0002002

2000-02-04

Abstract: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.

Artificial Intelligence
A Base Logic for Default Reasoning

Zhou Beihai,Mao Yi

DOI: https://doi.org/10.1007/s11466-006-0032-5

2006-01-01

Frontiers of Philosophy in China

Abstract:Based on a close study of benchmark examples in default reasoning, such as Nixon Diamond, Penguin Principle, etc., this paper provides an in depth analysis of the basic features of default reasoning. We formalize default inferences based on Modus Ponens for Default Implication, and mark the distinction between “local inferences” (to infer a conclusion from a subset of given premises) and “global inferences” (to infer a conclusion from the entire set of given premises). These conceptual analyses are captured by a formal semantics that is built upon the set-selection function technique. A minimal logic system M of default reasoning that accommodates Modus Ponens for Default Implication and suitable for local inferences is proposed, and its soundness is proved.
Are LLMs classical or nonmonotonic reasoners? Lessons from generics

Alina Leidinger,Robert van Rooij,Ekaterina Shutova

2024-06-12

Abstract:Recent scholarship on reasoning in LLMs has supplied evidence of impressive performance and flexible adaptation to machine generated or human feedback. Nonmonotonic reasoning, crucial to human cognition for navigating the real world, remains a challenging, yet understudied task. In this work, we study nonmonotonic reasoning capabilities of seven state-of-the-art LLMs in one abstract and one commonsense reasoning task featuring generics, such as 'Birds fly', and exceptions, 'Penguins don't fly' (see Fig. 1). While LLMs exhibit reasoning patterns in accordance with human nonmonotonic reasoning abilities, they fail to maintain stable beliefs on truth conditions of generics at the addition of supporting examples ('Owls fly') or unrelated information ('Lions have manes'). Our findings highlight pitfalls in attributing human reasoning behaviours to LLMs, as well as assessing general capabilities, while consistent reasoning remains elusive.

Computation and Language,Artificial Intelligence
Non-monotonic Reasoning and Argumentation System

XIE Kai-bo,LIAO Bei-shui

DOI: https://doi.org/10.3969/j.issn.1674-8425(s).2013.09.007

2013-01-01

Abstract:As first-order logic alone cannot depict the non-monotony in the human reasoning process,logicians try to establish new logic systems to deal with this problem,and these new systems are called non-monotonic logic.Traditionally,they adopt the way of consistency and unprovability.However,since the late 1980s and early 1990s,more and more logicians have begun to use a new way to realize the non-monotony of reasoning,which is argumentation.This paper will expound argumentation system and how argumentation framework is used to depict non-monotonic reasoning,and analyze the connections between them.
Complexity of Prioritized Default Logics

J. Rintanen

DOI: https://doi.org/10.1613/jair.554

2011-05-27

Abstract:In default reasoning, usually not all possible ways of resolving conflicts between default rules are acceptable. Criteria expressing acceptable ways of resolving the conflicts may be hardwired in the inference mechanism, for example specificity in inheritance reasoning can be handled this way, or they may be given abstractly as an ordering on the default rules. In this article we investigate formalizations of the latter approach in Reiter's default logic. Our goal is to analyze and compare the computational properties of three such formalizations in terms of their computational complexity: the prioritized default logics of Baader and Hollunder, and Brewka, and a prioritized default logic that is based on lexicographic comparison. The analysis locates the propositional variants of these logics on the second and third levels of the polynomial hierarchy, and identifies the boundary between tractable and intractable inference for restricted classes of prioritized default theories.

Artificial Intelligence
Restricted Semantics for Default Reasoning.

Zuoquan Lin,Chen

DOI: https://doi.org/10.1007/s10472-018-9592-z

IF: 1.019

2018-01-01

Annals of Mathematics and Artificial Intelligence

Abstract:In nonmonotonic reasoning there are the problems of inconsistency and incoherence in general, and in default reasoning there may be only one trivial extension or no extension in special. We propose the restricted semantics of four-valued logic for default reasoning to resolve the problems of inconsistency and incoherence and in the meantime retain classical extensions in the presence of consistency and coherency. The restricted semantics can maintain both the expressiveness and reasoning ability of default logic. We provide a transformation approach to compute the restricted extensions by reducing them into classical ones.
Strong Backdoors for Default Logic

Johannes K. Fichte,Arne Meier,Irena Schindler

DOI: https://doi.org/10.1145/3655024

2024-03-30

ACM Transactions on Computational Logic

Abstract:In this paper, we introduce a notion of backdoors to Reiter’s propositional default logic and study structural properties of it. Also we consider the problems of backdoor detection (parameterised by the solution size) as well as backdoor evaluation (parameterised by the size of the given backdoor), for various kinds of target classes (CNF, KROM, MONOTONE) and all SCHAEFER classes. Also, we study generalisations of HORN-formulas, namely, QHORN, RHORN, as well as DUALHORN. For these classes, we also classify the computational complexity of the implication problem. We show that backdoor detection is fixed-parameter tractable for the considered target classes, and prove a complete trichotomy for backdoor evaluation. The problems are either fixed-parameter tractable, para-DeltaP2-complete, or para-NP-complete, depending on the target class.

computer science, theory & methods,logic
Where Fail-Safe Default Logics Fail

Paolo Liberatore

DOI: https://doi.org/10.1145/1227839.1227842

2004-03-19

Abstract:Reiter's original definition of default logic allows for the application of a default that contradicts a previously applied one. We call failure this condition. The possibility of generating failures has been in the past considered as a semantical problem, and variants have been proposed to solve it. We show that it is instead a computational feature that is needed to encode some domains into default logic.

Artificial Intelligence,Logic in Computer Science
Iterated Revision In Default Logic

Shuang Ren,Guilin Qi,Zuoquan Lin

2009-01-01

Abstract:Default logic is supposed to reason with static and consistent default theory, i.e., it can not deal with inconsistencies arising in the situation that an agent receives a sequence of information represented by default theories. To overcome the problem, revision operators for revising default theories have been investigated by many researchers. But none of them has discussed the postulates which are desirable to be satisfied by the operators for revision of default theories, especially the operators for iterated revision. In this paper, we reformulate the existing postulates for belief revision and iterated belief revision. to default logic, and show that some of the reformulated postulates are rational to govern the process of revising default theories while others are not. A representation theorem for those reasonable postulates is provided.
A Demonic Outcome Logic for Randomized Nondeterminism

Noam Zilberstein,Dexter Kozen,Alexandra Silva,Joseph Tassarotti

2024-10-30

Abstract:Programs increasingly rely on randomization in applications such as cryptography and machine learning. Analyzing randomized programs has been a fruitful research direction, but there is a gap when programs also exploit nondeterminism (for concurrency, efficiency, or algorithmic design). In this paper, we introduce Demonic Outcome Logic for reasoning about programs that exploit both randomization and nondeterminism. The logic includes several novel features, such as reasoning about multiple executions in tandem and manipulating pre- and postconditions using familiar equational laws -- including the distributive law of probabilistic choices over nondeterministic ones. We also give rules for loops that both establish termination and quantify the distribution of final outcomes from a single premise. We illustrate the reasoning capabilities of Demonic Outcome Logic through several case studies, including the Monty Hall problem, an adversarial protocol for simulating fair coins, and a heuristic based probabilistic SAT solver.

Logic in Computer Science,Programming Languages
Weak Permission is not Well-Founded, Grounded and Stable

Guido Governatori

2024-11-16

Abstract:We consider the notion of weak permission as the failure to conclude that the opposite obligation. We investigate the issue from the point of non-monotonic reasoning, specifically logic programming and structured argumentation, and we show that it is not possible to capture weak permission in the presence of deontic conflicts under the well-founded, grounded and (sceptical) stable semantics.

Logic in Computer Science,Artificial Intelligence
On resolving conflicts between arguments

Nico Roos

DOI: https://doi.org/10.1111/0824-7935.00120

2022-09-21

Abstract:Argument systems are based on the idea that one can construct arguments for propositions; i.e., structured reasons justifying the belief in a proposition. Using defeasible rules, arguments need not be valid in all circumstances, therefore, it might be possible to construct an argument for a proposition as well as its negation. When arguments support conflicting propositions, one of the arguments must be defeated, which raises the question of \emph{which (sub-)arguments can be subject to defeat}? In legal argumentation, meta-rules determine the valid arguments by considering the last defeasible rule of each argument involved in a conflict. Since it is easier to evaluate arguments using their last rules, \emph{can a conflict be resolved by considering only the last defeasible rules of the arguments involved}? We propose a new argument system where, instead of deriving a defeat relation between arguments, \emph{undercutting-arguments} for the defeat of defeasible rules are constructed. This system allows us, (\textit{i}) to resolve conflicts (a generalization of rebutting arguments) using only the last rules of the arguments for inconsistencies, (\textit{ii}) to determine a set of valid (undefeated) arguments in linear time using an algorithm based on a JTMS, (\textit{iii}) to establish a relation with Default Logic, and (\textit{iv}) to prove closure properties such as \emph{cumulativity}. We also propose an extension of the argument system that enables \emph{reasoning by cases}.

Artificial Intelligence
Restricted Four-Valued Logic for Default Reasoning.

Chen Chen,Zuoquan Lin

DOI: https://doi.org/10.1007/978-3-319-25159-2_4

2015-01-01

Abstract:In Reiter's default logic, it is possible that no useful information can be brought from inconsistent knowledge or no extension of incoherent default theories exists. In this paper, based on Belnap's four-valued logic, we propose a new variant of default logic called the restricted four-valued default logic to tolerate inconsistency and incoherency of knowledge in default reasoning. Our proposal can maintain both the expressive power of full default logic and the ability of default reasoning. Moreover, we present a transformation-based approach to compute the restricted four-valued extensions.

Limitations of Skeptical Default Reasoning

A logic for default reasoning

Axiomatic Aspects of Default Inference

Defeasible deontic logic: arguing about permission and obligation

Going Forth and Drawing Back: an Intensional Approach in Nonmonotonic Inference.

Decision under Insufficient Evidence: A Scalable Probabilistic Way

Reasoning about Exceptions in Ontologies: from the Lexicographic Closure to the Skeptical Closure

Non-Monotonicity in Probabilistic Reasoning

Uniform semantic treatment of default and autoepistemic logics

A Base Logic for Default Reasoning

Are LLMs classical or nonmonotonic reasoners? Lessons from generics

Non-monotonic Reasoning and Argumentation System

Complexity of Prioritized Default Logics

Restricted Semantics for Default Reasoning.

Strong Backdoors for Default Logic

Where Fail-Safe Default Logics Fail

Iterated Revision In Default Logic

A Demonic Outcome Logic for Randomized Nondeterminism

Weak Permission is not Well-Founded, Grounded and Stable

On resolving conflicts between arguments

Restricted Four-Valued Logic for Default Reasoning.