VOICE DATA PROVIDING SYSTEM AND DEVICE AND PROGRAM FOR GENERATING VOICE DATA

摘要

PROBLEM TO BE SOLVED: To prevent generation of voice messages which include expressions being not appropriate as speech contents of celebrities when generating voice messages by speech synthesis using speech contents of voice messages designated by purchasers and voice feature data of celebrities. SOLUTION: A voice data providing system comprises; inappropriate expression dictionaries 251 to 25n where expressions being not appropriate as contents spoken by a specific speaker are described; a document analysis part 21 which performs document analysis of speech contents of a voice message designated by a purchaser and collates the document analysis result with inappropriate expression dictionaries to determine whether the voice message includes expressions being not inappropriate as speech contents of the speaker or not; and a speech synthesis processing part (a rhythm control part 23 and a voice waveform generation part 24) which performs the speech synthesis processing using the document analysis result and voice feature data of the speaker (rhythm data 261 to 26n and vocal quality data 271 to 27n for individual speakers) to output speech synthesis data in the case that it is determined by the document analysis part 21 that inappropriate expressions are not included. COPYRIGHT: (C)2004,JPO&NCIPI