Speech Computing in Thai Polysyllabic Words

Abstract

The problems of speech computing in Thai polysyllabic word is the mistaken segmentation. The position of vowel character in Thai written words is normally occured in many positions. It can be at initial, upper, lower or final position of the words. With this character, different from other languages in the world, makes so many problems for creating the programs. The separations of words would be incorrect especially in the polysyllabic word borrowed from Pali, Sanskrit, Khmer etc. Thai people have learned the pronunciation of this exceptional words in the school while the computer can not do so. Therefore in this project, we propose the linguistic knowledge to solve this segmentation problems by using the phonetic syllable analysis.