WMT19
Related Papers
Exploiting Monolingual Data at Scale for Neural Machine Translation
AbstractWhile target-side monolingual data has been proven to be very useful to improve neural machine translation (briefly, NMT) through back translation, source-side monolingual data is not well investigated. In this …...
Assessing Human-Parity in Machine Translation on the Segment Level
AbstractRecent machine translation shared tasks have shown top-performing systems to tie or in some cases even outperform human translation. Such conclusions about system and human performance are, however, based on …...
Fine-grained linguistic evaluation for state-of-the-art Machine Translation
AbstractThis paper describes a test suite submission providing detailed statistics of linguistic performance for the state-of-the-art German-English systems of the Fifth Conference of Machine Translation (WMT20). The analysis covers 107 …...
Counterfactual Data Augmentation for Neural Machine Translation
AbstractWe propose a data augmentation method for neural machine translation. It works by interpreting language models and phrasal alignment causally. Specifically, it creates augmented parallel translation corpora by generating (path-specific) …...
Netmarble AI Center’s WMT21 Automatic Post-Editing Shared Task Submission
AbstractThis paper describes Netmarble’s submission to WMT21 Automatic Post-Editing (APE) Shared Task for the English-German language pair. First, we propose a Curriculum Training Strategy in training stages. Facebook Fair’s WMT19 …...
Relational Memory-Augmented Language Models
AbstractWe present a memory-augmented approach to condition an autoregressive language model on a knowledge graph. We represent the graph as a collection of relation triples and retrieve relevant relations for …...
REUSE: REference-free UnSupervised Quality Estimation Metric
AbstractThis paper describes our submission to the WMT2022 shared metrics task. Our unsupervised metric estimates the translation quality at chunk-level and sentence-level. Source and target sentence chunks are retrieved by …...