Training a statistical parser on noisy data by filtering

作者: Jinjing Jiang , John Chen

DOI:

关键词: Speech recognitionTop-down parsingParser combinatorParsing expression grammarSimple LR parserNatural language processingRecursive descent parserParsingArtificial intelligenceGLR parserBottom-up parsingComputer science

摘要: A filtering or identifying approach is disclosed and applied to the task of unsupervised adaptation a parsing model selected domain. In particular, unannotated text data from domain parsed using first parser. subset then used train an improved training module which can be type that outputs usable by parser another

参考文章(52)
Simon Corston-Oliver, Michael Gamon, Robert C. Moore, Zhu Zhang, Eric Ringger, Sentence realization model for a natural language generation system ,(2002)
John Milanski, Srinivas Gutta, Daniel Pelletier, Hugo J. Strubbe, Larry J. Eshelman, User interface/entertainment device that simulates personal interaction and charges external database with relevant data ,(2000)
Liran Brener, Joel Ovil, Method and apparatus for language processing ,(2004)
Jeremy Huntley Wright, Giuseppe Riccardi, Dijana Petrovska-Delacretaz, Allen Louis Gorin, Method for generating morphemes ASAJ. ,vol. 116, pp. 1879- 1879 ,(2000)
Mark Edward Epstein, Robert Todd Ward, Sharon Barbara Jones, Integrated development tool for building a natural language understanding application ,(2003)