In current phrase-based Statistical Machine Translation systems, more training data is generally better than less. However, a larger data set eventually introduces a larger model ...
A modelling procedure consists of a sequence of modelling tasks. With the increasing size of models during the development, the relations among modelling objects and their correspo...
Yuan Ren, Jens Lemcke, Andreas Friesen, Tirdad Rah...
Model M, a novel class-based exponential language model, has been shown to significantly outperform word n-gram models in state-of-the-art machine translation and speech recognit...
We describe the Arabic broadcast transcription system elded by IBM in the GALE Phase 4 machine translation evaluation. Key advances over our Phase 3.5 system include improvements ...
Brian Kingsbury, Hagen Soltau, George Saon, Stephe...
Process improvement activities in small and medium size enterprises (SME) are challenging due to small number of personnel and projects, people have to perform in a variety of role...