A maximum entropy approach to nonmonotonic reasoning

Goldszmidt, Moisés; Morris, Paul; Pearl, Judea

doi:10.1109/34.204904

Cited by 81 publications

(83 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The maximum entropy approach of (Goldzsmidt et al, 1993) is in the probabilistic tradition of the 1-entailment of system Z. It follows a similar intuition as Lexicographic Closure concerning conflicting defaults.…”

Section: Conditional Logics Their Core Properties and Related Workmentioning

confidence: 98%

Adaptively Applying Modus Ponens in Conditional Logics of Normality

Straßer¹

2013

Trends in Logic

View full text Add to dashboard Cite

show abstract

Section: Conditional Logics Their Core Properties and Related Workmentioning

confidence: 98%

Adaptively Applying Modus Ponens in Conditional Logics of Normality

Straßer¹

2013

Trends in Logic

View full text Add to dashboard Cite

show abstract

“…Another approach to default reasoning is based on the principle of maximum entropy [10]. To describe how the maximum-entropy ranking of possible worlds can be computed, we need some additional terminology.…”

Section: Reasoning About Default Rules In System Pmentioning

confidence: 99%

“…This selection problem is closely related to the problem of defining the closure of a set of defaults, which has been widely studied in the field of non-monotonic reasoning [14,9,10]. In particular, several proposals to define this closure are based on constructing a specific probability distribution [4,10]. As we will show, it is possible to lift these approaches and thus obtain an efficient and principled way to construct MLNs that are compatible with a given set of defaults.…”

Section: Introductionmentioning

confidence: 99%

“…Constructing this MLN requires us to select a specific probability distribution that is compatible with the defaults. This selection problem is closely related to the problem of defining the closure of a set of defaults, which has been widely studied in the field of non-monotonic reasoning [14,9,10]. In particular, several proposals to define this closure are based on constructing a specific probability distribution [4,10].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Constructing Markov Logic Networks from First-Order Default Rules

Kuželka

Davis

Schockaert

2016

Inductive Logic Programming

View full text Add to dashboard Cite

Abstract. Expert knowledge can often be represented using default rules of the form "if A then typically B". In a probabilistic framework, such default rules can be seen as constraints on what should be derivable by MAP-inference. We exploit this idea for constructing a Markov logic network M from a set of first-order default rules D, such that MAP inference from M exactly corresponds to default reasoning from D, where we view first-order default rules as templates for the construction of propositional default rules. In particular, to construct appropriate Markov logic networks, we lift three standard methods for default reasoning. The resulting Markov logic networks could then be refined based on available training data. Our method thus offers a convenient way of using expert knowledge for constraining or guiding the process of learning Markov logic networks.

show abstract

“…It is equivalent to entailment in System logic-based entailment by Lamarre [37]. Finally, mainly to solve problems with property inheritance from classes to exceptional subclasses, the maximum entropy approach was proposed by Goldszmidt et al [30]; lexicographic entailment was introduced by Lehmann [39] and Benferhat et al [3]; conditional entailment was proposed by Geffner [25,26]; and a belief function approach was suggested by Benferhat et al [6].…”

Section: Introductionmentioning

confidence: 99%