AUTOMATIC GENERATING DEVICE FOR INITIALLY WRITTEN TEXT, VOICE RECOGNITION DEVICE, AND RECORDING MEDIUM

摘要

PROBLEM TO BE SOLVED: To automatically generate phoneme-notation text data with breaking symbols. SOLUTION: A morpheme analysis of initially written text data 101 having KANJI (Chinese character) and KANA (Japanese syllabary) mixed is made to divide the character string of the initially written KANJI-KANA mixed text data into morphemes. The divided character strings of the text data are converted by a pattern matching part 107 into readings (phoneme notation) according to a reading dictionary 102. When there are readings for one character string, phoneme notation matching the initially written phoneme-notation text data is determined among the readings. An insertion symbol is inserted into the phoneme-notation text data whose reading is determined at a division position with the divided text data.