PURPOSE: A device for automatically generating a voice using a text to speech function is provided to record listed words or phrases, and to automatically generate an announcing voice by using the text to speech function for words and phrases which are not recorded, thereby minimizing the usage of text to speech synthetic sounds to prevent deterioration of sound quality. CONSTITUTION: A voice storage(20) stores a voice of an announcer to give information to a user through a telephone network. An input/output device(30) records a changed announcing sentence again, and transmits the recorded sentence to the voice storage. A voice reproducer(40) reproduces the voice recorded in the voice storage. A text to speech unit(50) comprises as follows. A word similarity search engine(51) searches word similarity of the inputted sentence. A text to speech engine(52) converts the inputted sentence into a voice. A LAN interface(53) receives the changed sentence and a message, and transmits a synthetic voice to a system.
展开▼
