摘要 |
<P>PROBLEM TO BE SOLVED: To accurately extract a word that appears frequently in text, without using any dictionary. <P>SOLUTION: A word extraction device 1 includes: a target character string extraction unit for extracting, as target character strings from text, character strings that have a predetermined number of characters or more and appear in the text a predetermined number of times or more; a first deletion unit for deleting, from a group of the target character strings extracted by the target character string extraction unit, a character string that is a partial character string of other target character strings and appears at positions other than positions included in the other target character strings in the text a less number of times than the predetermined number of times; and a word extraction unit for setting, as a word, a character string included in the group of the target character strings after the deletion of the character string by the first deletion unit. <P>COPYRIGHT: (C)2013,JPO&INPIT |