Hassan Sawaf to Discuss the Benefits of Hybrid Machine Translation for Media Monitoring
During the Eighth Conference of the Association for Machine Translation in the Americas (AMTA), AppTek Chief Scientist Hassan Sawaf, Ph.D., will discuss best practices for implementing Hybrid Machine Translation to improve media monitoring translations in fluency, context and accuracy. His presentation entitled, "Hybrid Machine Translation Applied to Media Monitoring," will outline how a Hybrid Machine Translation approach delivers better results than a pure rule-based and a pure corpus-based approach for both written and spoken input. This presentation will also demonstrate how to increase language model quality for dialect language speech recognition by using non-dialect, non-spontaneous language resources.
Hassan Sawaf has over 10 years experience in the areas of Natural Language Processing, Machine Translation and Speech Recognition. He was project manager for Daimler Benz in Germany, senior researcher at the University of Aachen, where he did his graduate studies for his Diploma (1992-1998) and Ph.D. (1998-2003), and currently heads Research & Development for AppTek. Under his guidance as the head of R&D, AppTek recently introduced the first hybrid machine translation system that fully integrated two complete machine translation systems (statistical and rule-based) into one system that provides the performance and benefits of both approaches. His group also led the development of numerous automated speech recognition engines, which have been integrated into AppTek's numerous media monitoring systems.
Hilton Prince Kuhio Hotel
Friday, October 24, 2008 2:00 - 2:30pm