| 摘要 |
PURPOSE:To easily recognize a word to be an unknown word by a listener in the case of a wrong word by generating temporary reading information from a character plane concerning a word judged as the unknown word as the result of document analysis and adding a parameter to that word for changing velocity or volume. CONSTITUTION:An inputted sentence is analyzed by a sentence analyzing part 1 and in the case of the unknown word, temporary reading is given to the unknown word by a reading information preparation part 3 while referring to a dictionary 4 for reading information preparation. Further, the unknown word sounding parameter is given to the temporary reading of the unknown word by an unknown word sounding parameter setting part 5 so as to decelerate the sounding velocity 20%. When sounding processing is performed concerning all the words, the reading information preparation part 3 performs the link processing of reading information. On the other hand, the velocity or the volume is set concerning entire reading. Finally, a voice synthesizing part 6 synthesizes voices and outputs them based on the reading information. |
