| 摘要 |
<P>PROBLEM TO BE SOLVED: To solve the problem that the precision of selecting a word used in a translation is degraded because a regular arrangement has not been analyzed even if words and phrases are regularly arranged and there are some parts to be treated equivalently to a table and an itemized part in an original text. <P>SOLUTION: An area, in which one or more lines with character strings and characters equivalent to spaces to be continuous one or more are alternatively arranged, is extracted as a regular arrangement of the original text, and a set of words and phrases in the area is extracted. The set is collated with a set of pre-defined translation dictionaries, and selects an appropriate word used in the translation in response to the degree of checking the sets with one another. A reference range in a document to be used is appropriately restricted in just proportion, and the range is used for deciding the meaning and a part of speech and makes the best use of a constitutive property of the original text to form a translated document. <P>COPYRIGHT: (C)2008,JPO&INPIT |
