摘要 |
PROBLEM TO BE SOLVED: To provide a speech recognition system and a speech file recording system, which record input data as a speech file and automatically impart a file name capable of directly grasping its content. SOLUTION: A user term information extraction part 20 previously extracts user term information from information that a user inputted to an application 10 having usage other than speech recording usage and a speech recognition dictionary management part 30 adds vocabulary to a speech recognition dictionary 31 according to the user term information. Then the user inputs a speech through a speech input part 40 and a speech recognition part 50 recognizes the speech by using the speech recognition dictionary 31. A representative user term information selection part 60 extracts user term information included in the speech recognition result and selects one or more pieces of representative user term information from the user term information. A speech file recording part 70 records speech data as a speech file and gives a file name according to the representative user term information. COPYRIGHT: (C)2003,JPO
|