Real-Word Spelling Correction using GoogleWeb 1T n-gram with Backoff

摘要：

We present a method for correcting real-word spelling errors using the GoogleWeb 1T n-gram data set and a normalized and modied version of the Longest Common Subsequence (LCS) string matching algo- rithm. Our method is focused mainly on how to improve the correction recall (the fraction of er- rors corrected) while keeping the correction preci- sion (the fraction of suggestions that are correct) as high as possible. Evaluation results on a standard data set show that our method performs very well.

关键词： Real-word spelling correction Google web 1T n-gram

作者: Aminul ISLAM Diana INKPEN

作者单位: Department of Computer Science, SITE University of Ottawa Ottawa, ON, Canada

会议类型: 国际会议

会议名称: International Conference on Natural Language Processing and Knowledge Engineering(IEEE自然语言处理与知识工程国际会议 IEEE NLP-KE 2009)

会议地点: 大连

会议语种:英文

页码: 1-8

在线出版日期: 2009-09-24（万方平台首次上网日期，不代表论文的发表时间）