A
Alberto Poncelas
Researcher at Dublin City University
Publications - 51
Citations - 533
Alberto Poncelas is an academic researcher from Dublin City University. The author has contributed to research in topics: Machine translation & Test set. The author has an hindex of 11, co-authored 47 publications receiving 461 citations.
Papers
More filters
Investigating Backtranslation in Neural Machine Translation
Alberto Poncelas,Dimitar Shterionov,Andy Way,Gideon Maillette de Buy Wenniger,Peyman Passban +4 more
TL;DR: This paper investigated the effect of back-translated data on the performance of a neural machine translation (NMT) model for German-to-English translation and showed that back-translation has a significant impact on NMT performance.
Posted Content
Investigating Backtranslation in Neural Machine Translation
Alberto Poncelas,Dimitar Shterionov,Andy Way,Gideon Maillette de Buy Wenniger,Peyman Passban +4 more
TL;DR: This work investigates how using back-translated data as a training corpus -- both as a separate standalone dataset as well as combined with human-generated parallel data -- affects the performance of an NMT model.
SMT versus NMT: Preliminary comparisons for Irish
TL;DR: A preliminary comparison of statistical machine translation and neural machine translation for English→Irish in the fixed domain of public administration shows that while an out-of-the-box NMT system may not fare quite as well as the authors' tailor-made domain-specific SMT system, the future may still be promising for EN→GA NMT.
Proceedings ArticleDOI
Extracting In-domain Training Corpora for Neural Machine Translation Using Data Selection Methods
TL;DR: Three data selection approaches for machine translation systems are reviewed, namely Term Frequency– Inverse Document Frequency, Cross-Entropy Difference and Feature Decay Algorithm, and the results showed that using data selection also improved the performance, though the gain is not as much as for SMT systems.
Feature decay algorithms for neural machine translation
TL;DR: It is revealed that it is possible to find a subset of sentence pairs, that outperforms by 1.11 BLEU points the full training corpus, when used for training a German-English NMT system.