摘要 |
<P>PROBLEM TO BE SOLVED: To provide a sentence section extracting device capable of automatically extracting a sentence section including a regular expression from text data composed of a plurality of sentences, and to provide a program. <P>SOLUTION: A learning section 10 extracts tree structure and subtrees from learning data, generates a weak learning device on the subtrees, learns a weighting coefficient using the generated weak learning device, and generates final hypothetical information based on the weighting coefficient and the weak learning device. An extracting section 20 extracts a keyword and a key sentence from the text data composed of the plurality of sentences, and determines whether the regular expression is included in the sentence section based on the final hypothetical information to specify the sentence section. <P>COPYRIGHT: (C)2007,JPO&INPIT |