| 摘要 |
PURPOSE:To make it possible to automatically learn words other than the unification of translated words by automatically detecting the inconsistency of translated words between derivatives from a translation text. CONSTITUTION:A bilingual dictionary describing the information of derivatives, a stop word file storing words whose translated words are not uniformed and an exceptional word file storing derivatives whose translated words are not uniformed in each of derivatives are prepared and parts corresponding to words other than the stop words in an original text are extracted from a translation text and stored in a corresponding relation determining table (12). Derivatives other than exceptional words having inconsistent corresponding relation are extracted from the table in each derivative and exceptional words having inconsistent corresponding relation are extracted in each word (13) and displayed on a display screen (14), or the extracted words are added (15) to the stop word file and the exceptional word file. Consequently the detection accuracy of the inconsistency of translated words in the text obtained after post editing work can be improved and labor for the post editing work can be reduced. |
