Generic Techniques for Multilingual Speech Technology Applications
Julie Carson-Berndsen, Michael Walsh
Abstract : This paper is concerned with generic techniques for representing and evaluating phonological information in multilingual speech technology applications. A computational linguistic model of phonological interpretation is enhanced by a framework for constructing and evaluating phonotactic automata and by a generic lexicon model. The techniques make way for the extension of current speech technology to languages which have received little attention thus far.