TR-I-0035 :1988.08

Christian Boitet

Representation and computation of units of translation for Machine Interpretation of spoken texts

Abstract:This paper deals with an overall organization of the Machine Interpretation system, in which heterogeneous elements such as speech recognition, machine translation and speech synthesis should be put together in a unified framework (functional, computational and linguistic organization). Specifically a layered weighted lattice as the coarsest representation of a unit of translation is proposed in the spirit of the blackboard technique. This work was done when the author was a visiting research scientist of ATR Interpreting Telephony Research Laboratories.