| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a device and a program for dividing text data that can accurately section text data in phonemic description, even when the text data include a phoneme string where a plurality of consonants and a plurality of vowels are connected. <P>SOLUTION: The text data dividing device 1 generates a plurality of pieces of phoneme-string divided hypothesis data, by dividing the text data in phonemic description according to a previously registered phoneme string and selects and outputs minimum phoneme string divided hypothesis data having the smallest number of connection points as the number of places where phoneme strings are connected in the phoneme string divided hypothesis data, and is equipped with a hypothesis generating part 3, a phoneme string dictionary database 5, a character string dictionary database 7, a number-of-connection-points calculating part 9, a hypothesis deleting part 11, and a minimum divided hypothesis output part 13. <P>COPYRIGHT: (C)2005,JPO&NCIPI |
