摘要 |
<P>PROBLEM TO BE SOLVED: To provide a web site system for voice data transcription, which enables many and unspecified users to transcribe a voice in a moving image content or an audio content which cannot be temporarily stopped like live streaming, in real time in cooperation. <P>SOLUTION: A determination unit 22 acquires period audio data PAD distributed within a prescribed time period T which is a prescribed time period T before the input time of fragment text data, from an audio data storage unit 14 and determines whether an audio data part being an audio pattern acoustically matching a fragment audio pattern PAP exists in the period audio data PAD or not. A data substitution unit 23 substitutes a text data part of text data stored in a voice recognition result storage unit 15, which corresponds to a part being the audio pattern determined to acoustically match the fragment audio pattern PAP by the determination unit 22, by the fragment text data or word fragment text data. <P>COPYRIGHT: (C)2013,JPO&INPIT |