| 摘要 |
PROBLEM TO BE SOLVED: To provide a system for foreign language sentence interpretation support system that can obtain more natural translation results. SOLUTION: At the outset, specific non-translated words and phrases such as compound noun phrase or the like are extracted from a translation object document (in English and so on) by a natural language processing. Then, the whole object document is translated by converting part of non-translated words and phrases to katakana Japanese syllabary transcription without translating the part of non-translated words and phrases word by word. Thus, reading as the whole text of which the contents are known becomes easy by transcribing katakana, for example 'AKUSHON MANEJA' instead of strangely translating 'Action Manager' word by word. Also, since even a user who does not know the contents can read other parts on the assumption that they are written in special words and phrases rather than an obscure verbatim translation, he can go on reading without being confused. |
