A neural network based solution for automatic typing errors correction

Jun Li, Karim Ouazzane, Hassan Kazemian, Yanguo Jing, Richard Boyd

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Multiple algorithms have been developed to correct user's typing mistakes. However, an optimal solution is hardly identified among them. Moreover, these solutions rarely produce a single answer or share common results, and the answers may change with time and context. These motivated this research to synthesize some distinct word correction algorithms to produce an optimal prediction based on database updates and neural network learning. In this paper, three distinct typing correction algorithms are integrated as a pilot research with key factors such as Time Change, Context Change and User Feedback being considered. Experimental results show that the developed WLR model (i. e., word-list neural network ranking model) achieves the best results in comparison with Levenshtein distance, Metaphone. and Two-Gram prediction algorithms throughout all stages. It achieves 57.50% Ranking First Hitting Rate with samples of category one and a best Ranking First Hitting Rate of 74.69% within category four. Further work is recommended to extend the number of parameters and integrate more algorithms to achieve a higher level of accuracy.

Original languageEnglish
Pages (from-to)889-896
Number of pages8
JournalNeural Computing and Applications
Volume20
Issue number6
DOIs
Publication statusPublished - 1 Sep 2011

Fingerprint

Error correction
Neural networks
Feedback

Keywords

  • Jaro distance
  • Jaro-Winkler distance
  • Levenshtein distance
  • Metaphone
  • Neural network
  • Ranking First Hitting Rate
  • Word 2-gram
  • Word correction algorithms

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software

Cite this

A neural network based solution for automatic typing errors correction. / Li, Jun; Ouazzane, Karim; Kazemian, Hassan; Jing, Yanguo; Boyd, Richard.

In: Neural Computing and Applications, Vol. 20, No. 6, 01.09.2011, p. 889-896.

Research output: Contribution to journalArticle

Li, Jun ; Ouazzane, Karim ; Kazemian, Hassan ; Jing, Yanguo ; Boyd, Richard. / A neural network based solution for automatic typing errors correction. In: Neural Computing and Applications. 2011 ; Vol. 20, No. 6. pp. 889-896.
@article{fa65891f3540430b9b082baa899b3cbf,
title = "A neural network based solution for automatic typing errors correction",
abstract = "Multiple algorithms have been developed to correct user's typing mistakes. However, an optimal solution is hardly identified among them. Moreover, these solutions rarely produce a single answer or share common results, and the answers may change with time and context. These motivated this research to synthesize some distinct word correction algorithms to produce an optimal prediction based on database updates and neural network learning. In this paper, three distinct typing correction algorithms are integrated as a pilot research with key factors such as Time Change, Context Change and User Feedback being considered. Experimental results show that the developed WLR model (i. e., word-list neural network ranking model) achieves the best results in comparison with Levenshtein distance, Metaphone. and Two-Gram prediction algorithms throughout all stages. It achieves 57.50{\%} Ranking First Hitting Rate with samples of category one and a best Ranking First Hitting Rate of 74.69{\%} within category four. Further work is recommended to extend the number of parameters and integrate more algorithms to achieve a higher level of accuracy.",
keywords = "Jaro distance, Jaro-Winkler distance, Levenshtein distance, Metaphone, Neural network, Ranking First Hitting Rate, Word 2-gram, Word correction algorithms",
author = "Jun Li and Karim Ouazzane and Hassan Kazemian and Yanguo Jing and Richard Boyd",
year = "2011",
month = "9",
day = "1",
doi = "10.1007/s00521-010-0492-3",
language = "English",
volume = "20",
pages = "889--896",
journal = "Neural Computing and Applications",
issn = "0941-0643",
publisher = "Springer Verlag",
number = "6",

}

TY - JOUR

T1 - A neural network based solution for automatic typing errors correction

AU - Li, Jun

AU - Ouazzane, Karim

AU - Kazemian, Hassan

AU - Jing, Yanguo

AU - Boyd, Richard

PY - 2011/9/1

Y1 - 2011/9/1

N2 - Multiple algorithms have been developed to correct user's typing mistakes. However, an optimal solution is hardly identified among them. Moreover, these solutions rarely produce a single answer or share common results, and the answers may change with time and context. These motivated this research to synthesize some distinct word correction algorithms to produce an optimal prediction based on database updates and neural network learning. In this paper, three distinct typing correction algorithms are integrated as a pilot research with key factors such as Time Change, Context Change and User Feedback being considered. Experimental results show that the developed WLR model (i. e., word-list neural network ranking model) achieves the best results in comparison with Levenshtein distance, Metaphone. and Two-Gram prediction algorithms throughout all stages. It achieves 57.50% Ranking First Hitting Rate with samples of category one and a best Ranking First Hitting Rate of 74.69% within category four. Further work is recommended to extend the number of parameters and integrate more algorithms to achieve a higher level of accuracy.

AB - Multiple algorithms have been developed to correct user's typing mistakes. However, an optimal solution is hardly identified among them. Moreover, these solutions rarely produce a single answer or share common results, and the answers may change with time and context. These motivated this research to synthesize some distinct word correction algorithms to produce an optimal prediction based on database updates and neural network learning. In this paper, three distinct typing correction algorithms are integrated as a pilot research with key factors such as Time Change, Context Change and User Feedback being considered. Experimental results show that the developed WLR model (i. e., word-list neural network ranking model) achieves the best results in comparison with Levenshtein distance, Metaphone. and Two-Gram prediction algorithms throughout all stages. It achieves 57.50% Ranking First Hitting Rate with samples of category one and a best Ranking First Hitting Rate of 74.69% within category four. Further work is recommended to extend the number of parameters and integrate more algorithms to achieve a higher level of accuracy.

KW - Jaro distance

KW - Jaro-Winkler distance

KW - Levenshtein distance

KW - Metaphone

KW - Neural network

KW - Ranking First Hitting Rate

KW - Word 2-gram

KW - Word correction algorithms

UR - http://www.scopus.com/inward/record.url?scp=80051670322&partnerID=8YFLogxK

U2 - 10.1007/s00521-010-0492-3

DO - 10.1007/s00521-010-0492-3

M3 - Article

VL - 20

SP - 889

EP - 896

JO - Neural Computing and Applications

JF - Neural Computing and Applications

SN - 0941-0643

IS - 6

ER -