摘要 |
PURPOSE:To provide a text voice synthesizing device which can improve the word connecting accuracy and the comprehensiveness of the synthetic sound quality in the paragraph generation processing. CONSTITUTION:A device which synthesizes the voices from a text includes the part-of-speech information having no part-of-speech, the morpheme analysis S1 which adds the part-of-speech information to the text words, and the paragrapgh generation S2 which secures the connection between the adjacent words based on the part-of-speech information obtained through the analysis S1. In the generation S2, the connection of words is not carried out by the part-of-speech information of the no part-of-speech. The part-of-speech information on the no part-of-speech can be stored in a user dictionary 9 in the analysis S1. Then the part-of-speech information of the words in the generation S2 of paragraph can be stored in an internal buffer 6. |