A Statistical Model Based on the Three Head Words for Detecting Article Errors

Ryo NAGATA; Tatsuya IGUCHI; Fumito MASUI; Atsuo KAWAI; Naoki ISU

doi:10.1093/ietisy/e88-d.7.1700

A Statistical Model Based on the Three Head Words for Detecting Article Errors

Ryo NAGATA, Tatsuya IGUCHI, Fumito MASUI, Atsuo KAWAI, Naoki ISU

Full Text Views

0

Share
Cite this

Summary :

In this paper, we propose a statistical model for detecting article errors, which Japanese learners of English often make in English writing. It is based on the three head words--the verb head, the preposition, and the noun head. To overcome the data sparseness problem, we apply the backed-off estimate to it. Experiments show that its performance (F-measure=0.70) is better than that of other methods. Apart from the performance, it has two advantages: (i) Rules for detecting article errors are automatically generated as conditional probabilities once a corpus is given; (ii) Its recall and precision rates are adjustable.

Publication: IEICE TRANSACTIONS on Information Vol.E88-D No.7 pp.1700-1706

Publication Date: 2005/07/01

Publicized

Online ISSN

DOI: 10.1093/ietisy/e88-d.7.1700

Type of Manuscript: PAPER

Category: Educational Technology

Cite this

Copy

Ryo NAGATA, Tatsuya IGUCHI, Fumito MASUI, Atsuo KAWAI, Naoki ISU, "A Statistical Model Based on the Three Head Words for Detecting Article Errors" in IEICE TRANSACTIONS on Information, vol. E88-D, no. 7, pp. 1700-1706, July 2005, doi: 10.1093/ietisy/e88-d.7.1700.
Abstract: In this paper, we propose a statistical model for detecting article errors, which Japanese learners of English often make in English writing. It is based on the three head words--the verb head, the preposition, and the noun head. To overcome the data sparseness problem, we apply the backed-off estimate to it. Experiments show that its performance (F-measure=0.70) is better than that of other methods. Apart from the performance, it has two advantages: (i) Rules for detecting article errors are automatically generated as conditional probabilities once a corpus is given; (ii) Its recall and precision rates are adjustable.
URL: https://globals.ieice.org/en_transactions/information/10.1093/ietisy/e88-d.7.1700/_p

Copy

@ARTICLE{e88-d_7_1700,
author={Ryo NAGATA, Tatsuya IGUCHI, Fumito MASUI, Atsuo KAWAI, Naoki ISU, },
journal={IEICE TRANSACTIONS on Information},
title={A Statistical Model Based on the Three Head Words for Detecting Article Errors},
year={2005},
volume={E88-D},
number={7},
pages={1700-1706},
abstract={In this paper, we propose a statistical model for detecting article errors, which Japanese learners of English often make in English writing. It is based on the three head words--the verb head, the preposition, and the noun head. To overcome the data sparseness problem, we apply the backed-off estimate to it. Experiments show that its performance (F-measure=0.70) is better than that of other methods. Apart from the performance, it has two advantages: (i) Rules for detecting article errors are automatically generated as conditional probabilities once a corpus is given; (ii) Its recall and precision rates are adjustable.},
keywords={},
doi={10.1093/ietisy/e88-d.7.1700},
ISSN={},
month={July},}

Copy

TY - JOUR
TI - A Statistical Model Based on the Three Head Words for Detecting Article Errors
T2 - IEICE TRANSACTIONS on Information
SP - 1700
EP - 1706
AU - Ryo NAGATA
AU - Tatsuya IGUCHI
AU - Fumito MASUI
AU - Atsuo KAWAI
AU - Naoki ISU
PY - 2005
DO - 10.1093/ietisy/e88-d.7.1700
JO - IEICE TRANSACTIONS on Information
SN -
VL - E88-D
IS - 7
JA - IEICE TRANSACTIONS on Information
Y1 - July 2005
AB - In this paper, we propose a statistical model for detecting article errors, which Japanese learners of English often make in English writing. It is based on the three head words--the verb head, the preposition, and the noun head. To overcome the data sparseness problem, we apply the backed-off estimate to it. Experiments show that its performance (F-measure=0.70) is better than that of other methods. Apart from the performance, it has two advantages: (i) Rules for detecting article errors are automatically generated as conditional probabilities once a corpus is given; (ii) Its recall and precision rates are adjustable.
ER -