| 摘要 |
PROBLEM TO BE SOLVED: To obtain a correct translation result without requiring pre-editing by a user even if a document is written in a colloquial style including abbreviations. SOLUTION: An English text, which is inputted through an input part 10 is stored temporarily in a text storage area 41. Translation is performed normally by sentences, so the English text stored in the area 41 is sectioned into sentences and stored in an input buffer area 42. Then, when the user instructs that colloquial abbreviations are to be rewritten into correct notation, a CPU 30 performs retrieval from a colloquial rewriting dictionary 62 where proper notations corresponding to abbreviations used in colloquial documents are registered. When a corresponding character string is found, it is judged whether or not there are spaces right before and after it and when so, the abbreviation to be rewritten is regarded and rewritten into the proper notation, but when not corresponding character string is not rewritten. Then, a morpheme analysis is taken to divide the sentence into words. |
