# Domain, Translationese and Noise in Synthetic Data for Neural Machine Translation

arXiv: https://arxiv.org/abs/1911.03362 · HTML: https://mlsanity.com/paper/1911.03362/

## Key facts

- Cited by 26 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Nov 2019.

## Built on by (most cited first, 26 of 26)

- [Beyond Distillation: Task-level Mixture-of-Experts for Efficient Inference](https://mlsanity.com/paper/2110.03742/) (2021)
- [Facebook AI WMT21 News Translation Task Submission](https://mlsanity.com/paper/2108.03265/) (2021)
- [Domain Adaptation and Multi-Domain Adaptation for Neural Machine Translation: A Survey](https://mlsanity.com/paper/2104.06951/) (2021)
- [Rejuvenating Low-Frequency Words: Making the Most of Parallel Data in Non-Autoregressive Translation](https://mlsanity.com/paper/2106.00903/) (2021)
- [Causal Direction of Data Collection Matters: Implications of Causal and Anticausal Learning for NLP](https://mlsanity.com/paper/2110.03618/) (2021)
- [The Volctrans GLAT System: Non-autoregressive Translation Meets WMT21](https://mlsanity.com/paper/2109.11247/) (2021)
- [Exploiting Biased Models to De-bias Text: A Gender-Fair Rewriting Model](https://mlsanity.com/paper/2305.11140/) (2023)
- [Comparing Feature-Engineering and Feature-Learning Approaches for Multilingual Translationese Classification](https://mlsanity.com/paper/2109.07604/) (2021)
- [Understanding In-Context Machine Translation for Low-Resource Languages: A Case Study on Manchu](https://mlsanity.com/paper/2502.11862/) (2025)
- [On the Language Coverage Bias for Neural Machine Translation](https://mlsanity.com/paper/2106.03297/) (2021)
- [Domain-Specific Text Generation for Machine Translation](https://mlsanity.com/paper/2208.05909/) (2022)
- [Original or Translated? A Causal Analysis of the Impact of Translationese on Machine Translation Performance](https://mlsanity.com/paper/2205.02293/) (2022)
- [Leveraging Synthetic Audio Data for End-to-End Low-Resource Speech Translation](https://mlsanity.com/paper/2406.17363/) (2024)
- [Multilingual Pre-training with Language and Task Adaptation for Multilingual Text Style Transfer](https://mlsanity.com/paper/2203.08552/) (2022)
- [SegAugment: Maximizing the Utility of Speech Translation Data with Segmentation-based Augmentations](https://mlsanity.com/paper/2212.09699/) (2022)
- [Pretraining Language Models Using Translationese](https://mlsanity.com/paper/2403.13638/) (2024)
- [Target-Side Augmentation for Document-Level Machine Translation](https://mlsanity.com/paper/2305.04505/) (2023)
- [On Systematic Style Differences between Unsupervised and Supervised MT and an Application for High-Resource Machine Translation](https://mlsanity.com/paper/2106.15818/) (2021)
- [Language Modelling Approaches to Adaptive Machine Translation](https://mlsanity.com/paper/2401.14559/) (2024)
- [Script Normalization for Unconventional Writing of Under-Resourced Languages in Bilingual Communities](https://mlsanity.com/paper/2305.16407/) (2023)
- [Text Style Transfer Back-Translation](https://mlsanity.com/paper/2306.01318/) (2023)
- [Translation Asymmetry in LLMs as a Data Augmentation Factor: A Case Study for 6 Romansh Language Varieties](https://mlsanity.com/paper/2603.25489/) (2026)
- [Synthesizing Monolingual Data for Neural Machine Translation](https://mlsanity.com/paper/2101.12462/) (2021)
- [Leveraging Synthetic Targets for Machine Translation](https://mlsanity.com/paper/2305.06155/) (2023)
- [TopXGen: Topic-Diverse Parallel Data Generation for Low-Resource Machine Translation](https://mlsanity.com/paper/2508.08680/) (2025)
- [Improving Retrieval-Augmented Neural Machine Translation with Monolingual Data](https://mlsanity.com/paper/2504.21747/) (2025)

## Abstract (arXiv)

The quality of neural machine translation can be improved by leveraging additional monolingual resources to create synthetic training data. Source-side monolingual data can be (forward-)translated into the target language for self-training; target-side monolingual data can be back-translated. It has been widely reported that back-translation delivers superior results, but could this be due to artefacts in the test sets? We perform a case study using French-English news translation task and separate test sets based on their original languages. We show that forward translation delivers superior gains in terms of BLEU on sentences that were originally in the source language, complementing previous studies which show large improvements with back-translation on sentences that were originally in the target language. To better understand when and why forward and back-translation are effective, we study the role of domains, translationese, and noise. While translationese effects are well known to influence MT evaluation, we also find evidence that news data from different languages shows subtle domain differences, which is another explanation for varying performance on different portions of the test set. We perform additional low-resource experiments which demonstrate that forward translation is more sensitive to the quality of the initial translation system than back-translation, and tends to perform worse in low-resource settings.