
At present, the computer system application of speech research is the most interesting topic in technology sphere. Also Thai Computerized Voice Synthesis is one of the interesting subjects. Thai Computerized Voice Synthesis consists of three major parts: Word Segmentation in Thai Sentence, Phonetic symbols generator and Voice generator. This research presents the method of Word Segmentation. “Wordbucket Algorithm” will be implemented in order to segment the longest word compared with dictionary. Phonetics symbols generators is generated by MPT-networks which produces basic phoneme. And then phonetics symbols generated by MPT-network will be integrated into one Thai syllable.