| 摘要 |
Words uttered by a user are compared 2, with stored words 3; the word giving, the best "score" in the comparison is deemed to have been recognised. Where equal or similar scores occur the result is ambiguous and in that case a message is generated (eg by means of a speech synthesiser 7) containing a word for the user to confirm. If he does not, a second word may similarly be offered. <IMAGE> |
