Machine Translation based on Neural Network Language Models
Francisco Zamora-Mart · 2010
n-gramas Abstract: This paper describes a Machine Translation system that integrates a Neural Network Language Model in the decoding process. This work is motivated by the excellent performance of these connectionist language models. So far, the use of Neural Network Language Models in the translation systems is uncoupled: they are used in a second stage to rerank a N-best hypothesis list or to parse a word graph containing the N-best list. Our goal is to show the feasibility of using these language models within a fully integrated system.