Title: Statistical Learning of Languauge Pronunciation Structure

Authors: Filipp Korkmazskiy

Abstract: This paper presents a new approach to rule based pronunciation generation. A system presented in this paper can automatically learn a new language pronunciation structure and to use this knowledge for pronunciation generation for an arbitrary context sensitive language. Unlike conventional text-to-speech systems which are based on the cost expensive human expert knowledge about specific language, this system can learn by using only a set of spellings and pronunciations. The pronunciations can be obtained either from a pronunciation dictionary or from a phonetically labeled database. The system ability to learn pronunciation structure for any context sensitive language makes it a valuable tool for development of multilingual speech recognition systems. In this study we present experimental results on automatic generation of pronunciations for English, German, Spanish, French and Italian languages.

a01fk094.ps a01fk094.pdf