To obtain an easy-to-hear sentences by producing a sentence accordant with the decided treatment relation and meaning and outputting the sentence after converting it into the voices.
If the change of expression is decided when an electronic mail is read aloud, a sentence analysis part 207 analyzes the sentences of a specified electronic mail and produces a meaning expression. Then it's decided whether the sentences should be summarized. If the sentences are not summarized, the treatment relation between the transmitter and the receiver of the mail is decided with the ages, sexes of them. A sentence production part 208 produces the sentences based on the decided treatment relation and ages and sexes, and a sentence read-aloud part 205 synthesizes the voices in response to the produced sentences and reads them aloud. A voice transmission part 206 transmits the synthesized voices to a telephone set via a modem to end its processing. Thus, it's possible to produce the easy-to-hear sentences by reading aloud the produced sentences after changing them into the sentences where the treatment relation between the sentence transmitter and receiver is reflected.
YAGISAWA TSUYOSHI
FUJITA MINORU
