| 摘要 |
PROBLEM TO BE SOLVED: To let a speaker select the number of alternative words to be displayed on a correction window by providing a recognition component and a dictation editing component. SOLUTION: When the recognition component recognizes the word of 'make' erroneously as the word of 'fake', for example, the dictation editing component determines the present size of a rearrangeable correction window 103 and calculates the number of word displayable on that correction window 103. Next, the dictation editing component selects that number of words out of the list of recognized words (alternative words) having the highest probability. Then, these words are displayed on the correction window 103 while using standard window rearrangement rechnique. When the speaker further wants to watch more words from the list, it is enough only to rearrange the correction window 103 at the edge. |
