A Misspelling Intelligent Analysis Approach for Correcting Misspelled Words in English Text

This paper proposes an innovative MIA (Misspelling Intelligent Analysis) approach for efficient detection and intelligent correction of misspelled words. An integrity spelling correction approach is needed to consider both non-word errors and real-word errors. The MIA approach takes advantage of word frequency statistics, lexicon data, character distance and conditional probability for ranking suggestions of each misspelling having non-word errors. Drawing upon the context information, the overall score or probability is calculated and regarded as an access key for real-word errors correction in the MIA approach. Especially, features compensation and combination are provided so as to improve the accuracy of real-word errors correction in the articles of Chinese students. Finally, the experiments show that the MIA approach is capable of providing a better performance of error detection, discrimination and correction than current methods of dealing with misspelled words.