Six Challenges for Neural Machine Translation
Six Challenges for Neural Machine Translation
Philipp Koehn,Rebecca Knowles
2017 · DOI: 10.18653/v1/W17-3204
1,404 Citations
TLDR
This work explores six challenges for neural machine translation: domain mismatch, amount of training data, rare words, long sentences, word alignment, and beam search and shows both deficiencies and improvements over the quality of phrase-based statistical machine translation.
Abstract
We explore six challenges for neural machine translation: domain mismatch, amount of training data, rare words, long sentences, word alignment, and beam search. We show both deficiencies and improvements over the quality of phrase-based statistical machine translation.