KNGED: A tool for grammatical error diagnosis of Chinese sentences

Tao Hsing Chang, Yao Ting Sung, Jia Fei Hong, Jen I. Chang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Citations (Scopus)

Abstract

The main purpose of this paper is to propose a method that can automatically detect whether there are any grammatical errors as well as identify their error types. The framework of this method is based on a rule base to identify common grammatical errors. This rule base contains manually constructed rules and rules that are automatically machine generated. This paper further proposes algorithms which can apply these rules to determine whether a sentence is incorrect as well as what types of errors it belongs to. Experimental results show that the F1-measure of the proposed method is 0.64 and 0.30 on detection and identification, respectively.

Original languageEnglish
Title of host publicationWorkshop Proceedings of the 22nd International Conference on Computers in Education, ICCE 2014
EditorsThepchai Supnithi, Siu Cheung Kong, Ying-Tien Wu, Tomoko Kojiri, Chen-Chung Liu, Hiroaki Ogata, Akihiro Kashihara
PublisherAsia-Pacific Society for Computers in Education
Pages48-55
Number of pages8
ISBN (Electronic)9784990801427
Publication statusPublished - 2014
Event22nd International Conference on Computers in Education, ICCE 2014 - Nara, Japan
Duration: 2014 Nov 302014 Dec 4

Publication series

NameWorkshop Proceedings of the 22nd International Conference on Computers in Education, ICCE 2014

Other

Other22nd International Conference on Computers in Education, ICCE 2014
Country/TerritoryJapan
CityNara
Period2014/11/302014/12/04

Keywords

  • Automatic diagnosis
  • CFL
  • Chinese grammar
  • Grammatical error
  • KNGED
  • Rule-based method

ASJC Scopus subject areas

  • Computer Science Applications
  • Human-Computer Interaction
  • Education

Fingerprint

Dive into the research topics of 'KNGED: A tool for grammatical error diagnosis of Chinese sentences'. Together they form a unique fingerprint.

Cite this