Identification in the Limit of Substitutable Context-Free Languages

Clark, Alexander; Eyraud, Rémi

doi:10.1007/11564089_23

Cited by 52 publications

(93 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) The language can be learned from the entire class of computable generative probability distributions for language. Thus, the method is not restricted to particular classes of language structure, such as finite state or probabilistic context-free grammars, in contrast, to many important theoretical results (e.g., Angluin, 1988;Clark & Eyraud, 2007;Feldman et al, 1969). (2) The learner does not merely approximate the language, as in most positive probabilistic results, but identifies the generating distribution precisely.…”

Section: Gold Revisited: Generative Model Identification In the Limitmentioning

confidence: 99%

The probabilistic analysis of language acquisition: Theoretical, computational, and experimental analysis

2011

View full text Add to dashboard Cite

Keywords:Child language acquisition Poverty of the stimulus No negative evidence Bayesian models Minimum description length Simplicity principle Natural language Probabilistic models Identification in the limit a b s t r a c t There is much debate over the degree to which language learning is governed by innate language-specific biases, or acquired through cognition-general principles. Here we examine the probabilistic language acquisition hypothesis on three levels: We outline a novel theoretical result showing that it is possible to learn the exact generative model underlying a wide class of languages, purely from observing samples of the language. We then describe a recently proposed practical framework, which quantifies natural language learnability, allowing specific learnability predictions to be made for the first time. In previous work, this framework was used to make learnability predictions for a wide variety of linguistic constructions, for which learnability has been much debated. Here, we present a new experiment which tests these learnability predictions. We find that our experimental results support the possibility that these linguistic constructions are acquired probabilistically from cognition-general principles.

show abstract

Section: Gold Revisited: Generative Model Identification In the Limitmentioning

confidence: 99%

The probabilistic analysis of language acquisition: Theoretical, computational, and experimental analysis

2011

View full text Add to dashboard Cite

show abstract

“…Recently, Clark and Eyraud [4] presented a learning algorithm for a subclass of context-free languages which they called substitutable languages. Roughly speaking, substitutable languages are those context-free languages L which sat-isfy the condition that lur ∈ L if and only lvr ∈ L for pairs of strings u, v .…”

Section: Towards More Realistic Learning Scenariosmentioning

confidence: 99%

“…For the sake of better readability we recall the definition here in the form used by Clark and Eyraud [4]. Within this definition, L(R) denotes the languages described by representation R .…”

Section: Towards More Realistic Learning Scenariosmentioning

confidence: 99%

See 1 more Smart Citation

Inductive Inference and Language Learning

Zeugmann

2006

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…The first result along these lines was given by Clark and Eyraud [7]. They showed a polynomial identification in the limit result from positive data alone for a class of languages called the substitutable context free languages.…”

Section: Introductionmentioning

confidence: 97%

Polynomial Time Learning of Some Multiple Context-Free Languages with a Minimally Adequate Teacher

Yoshinaka

Clark

2012

Formal Grammar

Self Cite

View full text Add to dashboard Cite

Abstract. We present an algorithm for the inference of some Multiple Context-Free Grammars from Membership and Equivalence Queries, using the Minimally Adequate Teacher model of Angluin. This is an extension of the congruence based methods for learning some Context-Free Grammars proposed by Clark (ICGI 2010). We define the natural extension of the syntactic congruence to tuples of strings, and demonstrate we can efficiently learn the class of Multiple Context-Free Grammars where the non-terminals correspond to the congruence classes under this relation.

show abstract

Identification in the Limit of Substitutable Context-Free Languages

Cited by 52 publications

References 10 publications

The probabilistic analysis of language acquisition: Theoretical, computational, and experimental analysis

The probabilistic analysis of language acquisition: Theoretical, computational, and experimental analysis

Inductive Inference and Language Learning

Polynomial Time Learning of Some Multiple Context-Free Languages with a Minimally Adequate Teacher

Contact Info

Product

Resources

About