摘要 |
PROBLEM TO BE SOLVED: To solve problems of ambiguity in morpheme analysis (ambiguity of division into words and ambiguity of a reading of an identical-notation word). SOLUTION: A word dictionary 220 on a speech synthesis side 200 includes at least the vocabulary of a word dictionary 150 on a KANA(Japanese syllable)-to-KANJI(Chinese character) conversion side 100. Identifies (word ID) are assigned to registered vocabularies of both the dictionaries 150 and 220. The same word ID is assigned to identical words in the both 150 and 220. A word is determined through KANA-KANJI conversion on the KANA-KANJI conversion side 100 and then the word ID of the determined word is obtained. The obtained word ID is sent to the speech synthesis side 200. The speech synthesis side 200 performs retrieval from the dictionary 200 by using the word ID to obtain language information (notation, pronunciation, accent, part of speech, and accent combination information) on the word corresponding to the word ID. COPYRIGHT: (C)2004,JPO&NCIPI
|