摘要 |
PROBLEM TO BE SOLVED: To provide an information recognition device and an information recognizing program that can easily give a precise tag to archive data of a school lesson and a lecture based upon voice information when there is a text related to video, can precisely extract video parts corresponding to words, respective phrases, and a unit in the text, and enables a user to easily obtain a desired video part from long-time video data or a plurality of video data. SOLUTION: The device extracts words from the text as a material for school lesson and gives a tag to a video part by using appearance positions/orders/frequencies of the extracted words and the positional relation between word information extracted from writing information etc., recorded simultaneously with the school lesson or lecture and word information extracted from voice information. A voice recognition rate is improved by using information on the relationship to the school lesson/lecture to improve the precision of the tag to be given. COPYRIGHT: (C)2007,JPO&INPIT
|