Third ESCA/COCOSDA Workshop on Speech Synthesis

November 26-29, 1998
Jenolan Caves House, Blue Mountains, NSW, Australia

Multilingual Syllabification Using Weighted Finite-State Transducers

George Anton Kiraz, Bernd Möbius

Bell Labs, Lucent Technologies, Murray Hill, NJ, USA

This paper describes an approach to syllabification that has been incorporated into the English and German text-to-speech systems at Bell Labs. Implemented as a weighted finite-state transducer, the syllabifier is easily integrated - via mathematical composition - into the finite-state based text analysis component of the text-to-speech system. The weights are based on frequencies of onset, nucleus and coda types obtained from training data. While the training data is language-dependent, the formal approach is multilingual.


Full Paper

Bibliographic reference.  Kiraz, George Anton / Möbius, Bernd (1998): "Multilingual Syllabification Using Weighted Finite-State Transducers", In SSW3-1998, 71-76.