| 摘要 |
<P>PROBLEM TO BE SOLVED: To promptly perform voice reading of information desired by a user, by avoiding redundant reading. Ž<P>SOLUTION: A device for generating a voice reading sentence includes: a letter extraction unit 2 which accesses a form category data record unit 8 in which an appearing keyword in the form category and a priority corresponding to the keyword are recorded for each form category, and which extracts information for indicating a set of letter strings and a position of the letter string, from input information; a form category decision unit 3 for determining the form category by comparing the set of the letter strings, with the keyword in the form category data record unit 8; a letter string classification unit 4 for classifying the set of letter string into the letter string corresponding to each keyword included in the form category of the input information; and a reading sentence generation unit 6 for generating the text sentence for voice reading, by arranging the letter string corresponding to each keyword based on necessity of reading which is recorded corresponding to each keyword, or a priority order of reading. Ž<P>COPYRIGHT: (C)2010,JPO&INPIT Ž
|
