| 摘要 |
PROBLEM TO BE SOLVED: To prevent such a case where a sentence is divided at a reluctant part in a tag symbol and the translation becomes inadequate by inhibiting the division of an input sentence by means of a prescribed sentence separation symbol as long as the additional information is given to the separation symbol that is included in the input sentence. SOLUTION: When an input sentence consisting of the character data described in a 1st language and the additional information is inputted from an input part, a single-sentence segmentation part 2 checks the character of a scan position in an input document 1 contained in a memory or a file and then skips the scan until a tag end symbol appears if the present scan position is identical to a tag start symbol. Thereby, the sentence is segmented without being divided within a tag. In regard to the segmented character string, the relation between the additional information and the character data is stored at a tag separation part 3 and the additional information is separated from the character data. A translation part 4 translates the separated character data into a desired 2nd language and stores the correspondence of language equivalents. A tag restoration part 5 restores the tag into the translated equivalent of the 2nd language and outputs it. |
