Expanding the domain of a multi-lingual speech-to-speech translation system
Alon Lavie, Lori S. Levin, Puming Zhan, Maite Taboada, DONNA M. GATES, Mirella Lapata, Cortis Clark, Matthew Broadhead, Alex Waibel · 1997
JANUS is a multi-lingual speech-to-speech translation system, which has been designed to translate sponta-neous spoken language in a limited domain. In this paper, we describe our recent preliminary efforts to ex-pand the domain of coverage of the system from the rather limited Appointment Scheduling domain, to the much richer Travel Planning domain. We compare the two domains in terms of out-of-vocabulary rates and linguistic omplexity. We discuss the challenges that these differences impose on our translation system and some planned changes in the design of the system. Ini-tial evaluations on Travel Planning data are also pre-sented.