Computer Clusters

Training word alignment models on large cor- pora is a very time-consuming processes. This paper describes two parallel implementations of GIZA++ that accelerate this word align- ment process. One of …...