摘要 |
<p>PROBLEM TO BE SOLVED: To extract a daily word with high accuracy.SOLUTION: A history is extracted on each character string in a plurality of documents which are sent within an extraction target period prepared beforehand by history an extracting part 134 and which have undergone morphological analysis. Based on an identification model stored in an identification model storage part 136 and the history extracted by a daily word extracting part 138, it is determined whether each of the character strings in the documents having undergone the morphological analysis is a daily word, and it is extracted as a daily word.</p> |