On the inference of canonical context-free grammars

Fass, Leona F.

doi:10.1145/8307.8310

Cited by 4 publications

(3 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meera Blattner encouraged the SIGACT publication of (Fass 1983) and the late E.S. (Stewart) Bainbridge encouraged us to publish (Fass 1986). This all led to our interactions with an interdisciplinary research community, some of whose research is described above.…”

Section: Acknowledgmentsmentioning

confidence: 99%

“…We imposed formal constraints on language, considering those describable as context-free, and developed grammatical inference techniques to determine perfect, learnable language models. Based on the suggestion of Leon Levy and Aravind Joshi we represented language components structurally, which enabled us to obtain our original results [e.g., (Fass 1983(Fass , 1986]. For the structured language we proved existence of perfect generative and recognitive models; refined techniques to construct them effectively; and established procedures to infer these models from appropriate linguistic data, inductively.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Language, mathematics and other dangerous things

Fass

2004

SIGACT News

View full text Add to dashboard Cite

We describe our original automata-theoretic, highly mathematical perspective on language representation and learning, and compare with some current research advances and some that are proposed. We find benefits of a mathematical approach to "real life" language learning by humans and today's machines, and review work of researchers who adapt such computational techniques to complex system development and language-related machine processes. Examples include design of communicating agents and proposed development of the Semantic Web. While emphasizing the success of a mathematical perspective on language representation and learning, we also note the danger of assuming that "perfect" results will always be produced by a mathematical approach.

show abstract

Section: Acknowledgmentsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Language, mathematics and other dangerous things

Fass

2004

SIGACT News

View full text Add to dashboard Cite

show abstract

“…Every constituent phrase of L, as defined by the generative grammar G, is the yield of at least one skeletal subtree. When L is represented by S, the language can be described in terms of inductive construction of sentence shapes from phrase-structures: the skeletal subtrees of S and their yields [4,5].…”

Section: Defining 1vi Andmentioning

confidence: 99%

Learnability of CFLs: inferring syntactic models from constituent structure

Fass

1989

SIGART Bull.

View full text Add to dashboard Cite

Inductive inference is a learning process based on discovering models for bodies of knowledge, given sample information. The inference process we discuss here is concerned with inductive acquisition of syntactic models for context-free languages (CFLs), given appropriate language samples. The knowledge to be modeled in this case is any CFL L, with the model to be determined a recognitive or generative characterization of L's syntactic structure. L will be learned syntactically once a machine M recognizing L, or a context-free grammar (CFG) G generating L, is inductively inferred from a sentence sample. The capability of distinguishing between L and its complement, or of generating all and only L 's sentences, is the knowledge acquired, with the learner (inference process) gaining this knowledge by acquiring M or G. An observer (informant, teacher, or oracle) has such knowledge of L and can provide the learner with appropriate sample information to ensure that M or G is correctly identified.

show abstract

Meeting of the Association for Symbolic Logic, New York City, May 1987

Davis¹,

Parikh²

1988

J. symb. log.

View full text Add to dashboard Cite

On the inference of canonical context-free grammars

Cited by 4 publications

References 4 publications

Language, mathematics and other dangerous things

Language, mathematics and other dangerous things

Learnability of CFLs: inferring syntactic models from constituent structure

Meeting of the Association for Symbolic Logic, New York City, May 1987

Contact Info

Product

Resources

About