Some important issues on text analysis in a Chinese to Taiwanese TTS system

This paper deals with some important issues of a Chinese to Taiwanese TTS system. The first issue is to determine the meanings and pronunciations of the Chinese word for we in Chinese texts. There are two pronunciations of the word in Taiwanese. The two pronunciations are /ghun/ and /lan/ with the c...

Full description

Saved in:
Bibliographic Details
Published in2005 9th International Workshop on Cellular Neural Networks and Their Applications pp. 186 - 189
Main Authors Yih-Jeng Lin, Wang-Song Ji, Ming-Shing Yu, Shang-De Li
Format Conference Proceeding
LanguageEnglish
Published IEEE 2005
Subjects
Computer science
Educational institutions
Information management
Mathematics
Natural languages
Speech processing
Speech synthesis
Testing
Text analysis
Writing
Online AccessGet full text

Cover

More Information
Summary:This paper deals with some important issues of a Chinese to Taiwanese TTS system. The first issue is to determine the meanings and pronunciations of the Chinese word for we in Chinese texts. There are two pronunciations of the word in Taiwanese. The two pronunciations are /ghun/ and /lan/ with the corresponding meanings of the Chinese words. The second issue we bring up is to determine the prosodic words in a Taiwanese TTS system. The tone sandhi in Taiwanese is more common than that in a Mandarin TTS system. We try to find a method to determine the prosodic words of an input text. The achieved precision rate of the first issue we got is 81.4% for outside test, and the precision rate of the second issue is 79.5%.
ISBN:0780391853
9780780391857
ISSN:2165-0144
2165-0152
DOI:10.1109/CNNA.2005.1543192