In long-term translation projects, like Parliamentary text, there is a desire to build machine translation systems that can adapt to changes over time. We implement and examine a simple approach to continual learning for neural machine translation, exploring tradeoffs between consistency, the model’s ability to learn from incoming data, and the time a client would need to wait to obtain a newly trained translation system.

The MSLC (Metric Score Landscape Challenge) submissions for English-German, English-Spanish, and Japanese-Chinese are constrained systems built using Transformer models for the purpose of better evaluating metric performance in the WMT24 Metrics Task. They are intended to be representative of the performance of systems that can be built relatively simply using constrained data and with minimal modifications to the translation training pipeline.

In this second edition of the Metric Score Landscape Challenge (MSLC), we examine how automatic metrics for machine translation perform on a wide variety of machine translation output, ranging from very low quality systems to the types of high-quality systems submitted to the General MT shared task at WMT. We also explore metric results on specific types of data, such as empty strings, wrong- or mixed-language text, and more. We raise several alarms about inconsistencies in metric scores, some of which can be resolved by increasingly explicit instructions for metric use, while others highlight technical flaws.

Novel terminology and changes in terminology are often a challenge for machine translation systems. The passing of Queen Elizabeth II and the accession of King Charles III provide a striking example of translation shift in the real world, particularly in translation contexts that have ambiguity. Examining translation between French and English, we present a focused case-study of translations about King Charles III as produced both by publicly-available MT systems and by a neural machine translation system trained specifically on Canadian parliamentary text. We find that even in cases where human translators would have adequate context to disambiguate terms from the source language, machine translation systems do not always produce the expected output. Where we are able to analyze the training data, we note that this may represent artifacts in the data, raising important questions about machine translation updates in light of real world events.

Incorporating terminology into a neural machine translation (NMT) system is a feature of interest for many users of machine translation. In this case study of English-French Canadian Parliamentary text, we examine the performance of standard NMT systems at handling terminology and consider the tradeoffs between potential performance improvements and the efforts required to maintain terminological resources specifically for NMT.

The Metric Score Landscape Challenge (MSLC23) dataset aims to gain insight into metric scores on a broader/wider landscape of machine translation (MT) quality. It provides a collection of low- to medium-quality MT output on the WMT23 general task test set. Together with the high quality systems submitted to the general task, this will enable better interpretation of metric scores across a range of different levels of translation quality. With this wider range of MT quality, we also visualize and analyze metric characteristics beyond just correlation.

We revisit the topic of translation direction in the data used for training neural machine translation systems and focusing on a real-world scenario with known translation direction and imbalances in translation direction: the Canadian Hansard. According to automatic metrics and we observe that using parallel data that was produced in the “matching” translation direction (Authentic source and translationese target) improves translation quality. In cases of data imbalance in terms of translation direction and we find that tagging of translation direction can close the performance gap. We perform a human evaluation that differs slightly from the automatic metrics and but nevertheless confirms that for this French-English dataset that is known to contain high-quality translations and authentic or tagged mixed source improves over translationese source for training.

We describe the NRC-CNRC systems submitted to the AmericasNLP shared task on machine translation. We submitted systems translating from Spanish into Wixárika, Nahuatl, Rarámuri, and Guaraní. Our best neural machine translation systems used multilingual pretraining, ensembling, finetuning, training on parts of the development data, and subword regularization. We also submitted translation memory systems as a strong baseline.

We describe our neural machine translation systems for the 2021 shared task on Unsupervised and Very Low Resource Supervised MT, translating between Upper Sorbian and German (low-resource) and between Lower Sorbian and German (unsupervised). The systems incorporated data filtering, backtranslation, BPE-dropout, ensembling, and transfer learning from high(er)-resource languages. As measured by automatic metrics, our systems showed strong performance, consistently placing first or tied for first across most metrics and translation directions.

We describe the National Research Council of Canada (NRC) submissions for the 2020 Inuktitut-English shared task on news translation at the Fifth Conference on Machine Translation (WMT20). Our submissions consist of ensembled domain-specific finetuned transformer models, trained using the Nunavut Hansard and news data and, in the case of Inuktitut-English, backtranslated news and parliamentary data. In this work we explore challenges related to the relatively small amount of parallel data, morphological complexity, and domain shifts.

We present a study on using YiSi-2 with massive multilingual pretrained language models for machine translation (MT) reference-less evaluation. Aiming at finding better semantic representation for semantic MT evaluation, we first test YiSi-2 with contextual embed- dings extracted from different layers of two different pretrained models, multilingual BERT and XLM-RoBERTa. We also experiment with learning bilingual mappings that trans- form the vector subspace of the source language to be closer to that of the target language in the pretrained model to obtain more accurate cross-lingual semantic similarity representations. Our results show that YiSi-2’s correlation with human direct assessment on translation quality is greatly improved by replacing multilingual BERT with XLM-RoBERTa and projecting the source embeddings into the tar- get embedding space using a cross-lingual lin- ear projection (CLP) matrix learnt from a small development set.

We describe the National Research Council of Canada (NRC) neural machine translation systems for the German-Upper Sorbian supervised track of the 2020 shared task on Unsupervised MT and Very Low Resource Supervised MT. Our models are ensembles of Transformer models, built using combinations of BPE-dropout, lexical modifications, and backtranslation.

The Inuktitut language, a member of the Inuit-Yupik-Unangan language family, is spoken across Arctic Canada and noted for its morphological complexity. It is an official language of two territories, Nunavut and the Northwest Territories, and has recognition in additional regions. This paper describes a newly released sentence-aligned Inuktitut–English corpus based on the proceedings of the Legislative Assembly of Nunavut, covering sessions from April 1999 to June 2017. With approximately 1.3 million aligned sentence pairs, this is, to our knowledge, the largest parallel corpus of a polysynthetic language or an Indigenous language of the Americas released to date. The paper describes the alignment methodology used, the evaluation of the alignments, and preliminary experiments on statistical and neural machine translation (SMT and NMT) between Inuktitut and English, in both directions.

We describe the neural machine translation (NMT) system developed at the National Research Council of Canada (NRC) for the Kazakh-English news translation task of the Fourth Conference on Machine Translation (WMT19). Our submission is a multi-source NMT taking both the original Kazakh sentence and its Russian translation as input for translating into English.

The WMT18 shared task on parallel corpus filtering (Koehn et al., 2018b) challenged teams to score sentence pairs from a large high-recall, low-precision web-scraped parallel corpus (Koehn et al., 2018a). Participants could use existing sample corpora (e.g. past WMT data) as a supervisory signal to learn what a “clean” corpus looks like. However, in lower-resource situations it often happens that the target corpus of the language is the only sample of parallel text in that language. We therefore made several unsupervised entries, setting ourselves an additional constraint that we not utilize the additional clean parallel corpora. One such entry fairly consistently scored in the top ten systems in the 100M-word conditions, and for one task—translating the European Medicines Agency corpus (Tiedemann, 2009)—scored among the best systems even in the 10M-word conditions.

We present our semantic textual similarity approach in filtering a noisy web crawled parallel corpus using YiSi—a novel semantic machine translation evaluation metric. The systems mainly based on this supervised approach perform well in the WMT18 Parallel Corpus Filtering shared task (4th place in 100-million-word evaluation, 8th place in 10-million-word evaluation, and 6th place overall, out of 48 submissions). In fact, our best performing system—NRC-yisi-bicov is one of the only four submissions ranked top 10 in both evaluations. Our submitted systems also include some initial filtering steps for scaling down the size of the test corpus and a final redundancy removal step for better semantic and token coverage of the filtered corpus. In this paper, we also describe our unsuccessful attempt in automatically synthesizing a noisy parallel development corpus for tuning the weights to combine different parallelism and fluency features.

In this paper, we propose a new domain adaptation technique for neural machine translation called cost weighting, which is appropriate for adaptation scenarios in which a small in-domain data set and a large general-domain data set are available. Cost weighting incorporates a domain classifier into the neural machine translation training algorithm, using features derived from the encoder representation in order to distinguish in-domain from out-of-domain data. Classifier probabilities are used to weight sentences according to their domain similarity when updating the parameters of the neural translation model. We compare cost weighting to two traditional domain adaptation techniques developed for statistical machine translation: data selection and sub-corpus weighting. Experiments on two large-data tasks show that both the traditional techniques and our novel proposal lead to significant gains, with cost weighting outperforming the traditional methods.