DMCA
Learning Context-Free Grammars with a Simplicity Bias (2000)
Cached
Download Links
- [www.isle.org]
- [cll.stanford.edu]
- [www.isle.org]
- DBLP
Other Repositories/Bibliography
| Venue: | Proceedings of the Eleventh European Conference on Machine Learning |
| Citations: | 46 - 4 self |
BibTeX
@INPROCEEDINGS{Langley00learningcontext-free,
author = {Pat Langley and Sean Stromsten},
title = {Learning Context-Free Grammars with a Simplicity Bias},
booktitle = {Proceedings of the Eleventh European Conference on Machine Learning},
year = {2000},
pages = {220--228},
publisher = {Springer-Verlag}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
. We examine the role of simplicity in directing the induction of context-free grammars from sample sentences. We present a rational reconstruction of Wolff's SNPR -- the Grids system -- which incorporates a bias toward grammars that minimize description length. The algorithm alternates between merging existing nonterminal symbols and creating new symbols, using a beam search to move from complex to simpler grammars. Experiments suggest that this approach can induce accurate grammars and that it scales reasonably to more difficult domains. 1 Introduction In this paper we focus on the task of inducing context-free grammars from training sentences. Much recent work on this topic has dealt with learning finite-state structures, but there is considerable evidence that human language involves more powerful grammatical representations. In context-free grammar induction, the learner must find not only a set of grammatical rewrite rules but also the nonterminal symbols used in those r...
Keyphrases
context-free grammar simplicity bias nonterminal symbol beam search new symbol finite-state structure human language powerful grammatical representation description length considerable evidence sample sentence context-free grammar induction accurate grammar grid system rational reconstruction much recent work grammatical rewrite rule difficult domain
