Grammatical Error Correction With Alternating Structure Optimization
Tou Ng NUS Graduate School for Integrative Sciences and Engineering and National University of Singapore Published in: ·Proceeding HLT '11 Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies - Volume 1 Pages 915-923 Association for Computational Linguistics Stroudsburg, PA, USA ©2011 tableofcontents ISBN: 978-1-932432-87-9 2011 Article Bibliometrics ·Downloads (6 Weeks): 6 ·Downloads (12 Months): 66 ·Downloads (cumulative): 373 ·Citation Count: 12 Recent authors with related interests Concepts in this article powered by Concepts inGrammatical error correction with alternating structure optimization Linguistic prescription In linguistics, prescription denotes normative practices on such aspects of language use as spelling, grammar, pronunciation, and syntax. It includes judgments on what usages are socially proper and politically correct. Its aims may be to establish a standard language, to teach what is perceived within a particular society to be correct forms of language, or to advise on effective communication. morefromWikipedia Error detection and correction In information theory and coding theory with applications in computer science and telecommunication, error detection and correction or error control are techniques that enable reliable delivery of digital data over unreliable communication channels. Many communication channels are subject to channel noise, and thus errors may be introduced during transmission from the source to a receiver. morefromWikipedia Preposition and postposition Prepositions (or more generally, adpositions, see below) are a grammatically distinct class of words whose most central members characteristically express spatial relations (such as the English words in, under, toward) or serve to mark various syntactic functions and semantic roles (such as the English words of, for). morefromWikipedia Grammar checker A grammar checker, in computing terms, is a program, or part of a program, that attempts to verify written text for grammatical correctness. Grammar checkers are most often implemented as a feature of a larger program, such as a word processor, but are al
Grammatical Error Correction with Alternating Structure Optimization Inventor Name: Daniel Hermann Richard DAHLMEIER Name: Hwee Tou NG Contact Name: Jose Luis Rojas Roman Title: Manager Email: ilojlrr@nus.edu.sg Information Organization Name National University of Singapore Institutional ID Number 11208N Technology page URL http://ilo.technologypublisher.com/technology/12002 Detailed Technology Description ILO Ref: 11208N
Technology Overview Automatic correction of grammatical errors can be very useful for programs that help people for whom English is a second language. Most commonly, programs will use annotated learner texts to help with correcting grammatical errors. http://dl.acm.org/citation.cfm?id=2002588 However, these texts are few in number as compared to the millions of non- learner texts readily available. Also, they are expensive. Thus a system that can make use of non- learner texts as well is highly desirable. This method builds on Alternating Structure Optimization (ASO). The classifier uses a formula to predict a particular word (say a preposition) that can http://gtp.patsnap.com/technology/view/30878 be used at that position. It then computes how far off that prediction is from a correct word using an annotated or tagged text. Using this value it assigns each prediction a score. The prediction with the highest score is suggested as the correction. If the correction is same as the word used, then it makes no change. If there is a mismatch, the word is replaced with the correct one. Also, ASO is a multi-task learning algorithm that can learn from several related cases. A multi task algorithm can handle several related cases at the same time. Using a matrix algorithm, several words can be tested ar the same time. For example the program can learn to correct 8 or 10 different prepositions using the same learner text. In fact, this means non learner texts can also be used by the program for learning. By removing a particular word or phrase from an otherwise grammatically correct non learner text, the program can be taught to detect its misuse. In the same way as before, multi- task learning is possible in this scenario as well.Optimization. Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Techologies, pages grammatical error 915–923, Portland, Oregon, USA. Association for Computational Linguistics. http://www.aclweb.org/anthology/P11-1092. BibTeX entry: @inproceedings{dahlmeier-ng:2011:ACL-HLT2011, author = {Dahlmeier, Daniel and Ng, Hwee Tou}, title = {Grammatical Error Correction with grammatical error correction Alternating Structure Optimization}, booktitle = {Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Techologies}, month = {June}, year = {2011}, address = {Portland, Oregon, USA}, publisher = {Association for Computational Linguistics}, pages = {915--923}, url = {http://www.aclweb.org/anthology/P11-1092} } Further information on this paper: ACL Anthology ID: P11-1092 PDF fulltext Searchbench Document View Searchbench Citation Browser Searchbench Citation Contexts Searchbench Bibliographic Metadata ACL Anthology Network (beta) New ACL Anthology (beta)
be down. Please try the request again. Your cache administrator is webmaster. Generated Sat, 15 Oct 2016 20:59:38 GMT by s_ac4 (squid/3.5.20)