Natural Language Processing Of Chinese Text Automatic Proofreading

With the E-journal coming forth, it is more and more important toresolve the automatic detection and correction of Chinese text.Researching the automatic detection and correction of Chinese textbecomes an urgency task.So on the base of researching and analyzing the technology andmethod of the detection and correction of Chinese text, an improvedmethod is presented by the paper.Firstly, for the automatic detection and correction of punctuation, onthe base of splitting word and labeling part of speech of words, the paperpresents an arithmetic that drove by rules and information of context.And the results show that the arithmetic can resolve a majority ofpunctuation mistakes.Secondly, for the detection of word mistakes, after researching andanalyzing a variety of different methods presented by academic animproved method on the base of error-window technology that supportedby a large-scale contemporary Chinese corpus is presented. The methoddecreases the complexity of the traditional arithmetic by using theerror-window technology.Thirdly, for the correction of word mistakes, utilizing the characterof the word mistakes, an improved method on the base of traditionalmethods is presented. The method can give right correct suggestion for amajority of words with different mistakes by construct a confusingwords dictionary and a confusing characters aggregate. In the last part ofthe paper, a work plan for future to improve the percentage ofcorrectness of detecting and correcting the Chinese text is made out.