Automatic Syllabification of Bengali in SPPAS

This paper describes the automatic syllabification process of Bengali speech in SPPAS software. The process of detecting syllable boundaries has been carried out using a rule-based system applied to the permissible syllable structures of Bengali. Syllable structures of Bengali has undergone a fair amount of research. In this paper, we verify the existing syllabification rules of Bengali through a speech corpus and then describe its implementation through the resource toolkit of SPPAS software. Finally, by evaluating the automated syllabification system with the manually aligned syllables of the corpus, we have successfully implemented this automatization task, which can significantly help to syllabify large scale speech and text databases in Bengali.