| 摘要 |
PURPOSE: To improve efficiency for document correction by preferentially displaying a certain candidate by correcting not only spelling errors but also misunderstanding such as English words coined in Japan or inflection errors in the case of checking the error of a document, and excluding understanding difficulty caused by the mixture of plural errors. CONSTITUTION: A sentence segmenting part segments an inputted sentence for the unit of word. To the input, plural error judging modules 4 respectively generate candidates while using original methods. A judging module applying part 3 refers to a dictionary 6 while using the segmented word and applies any error judging module 4 corresponding to fixed order and stop conditions concerning the word not registered on the dictionary, and the result is stored. A display part displays the result provided by the judging module applying part 3. |
