# Analyzing the Use of Character-Level Translation with Sparse and Noisy Datasets

arXiv: https://arxiv.org/abs/2109.13723 · HTML: https://mlsanity.com/paper/2109.13723/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 2 of its 44 references that are in the map.
- First posted on arXiv 27 Sep 2021.

## Builds on (most cited first, 2 of 2)

- [Improving Statistical Machine Translation for a Resource-Poor Language Using Related Resource-Rich Languages](https://mlsanity.com/paper/1401.6876/) (2014)
- [A Hybrid Morpheme-Word Representation for Machine Translation of Morphologically Rich Languages](https://mlsanity.com/paper/1911.08117/) (2019)

## Built on by (most cited first, 2 of 2)

- [Are Character-level Translations Worth the Wait? Comparing ByT5 and mT5 for Machine Translation](https://mlsanity.com/paper/2302.14220/) (2023)
- [Automatic Correction of Writing Anomalies in Hausa Texts](https://mlsanity.com/paper/2506.03820/) (2025)

## Abstract (arXiv)

This paper provides an analysis of character-level machine translation models used in pivot-based translation when applied to sparse and noisy datasets, such as crowdsourced movie subtitles. In our experiments, we find that such character-level models cut the number of untranslated words by over 40% and are especially competitive (improvements of 2-3 BLEU points) in the case of limited training data. We explore the impact of character alignment, phrase table filtering, bitext size and the choice of pivot language on translation quality. We further compare cascaded translation models to the use of synthetic training data via multiple pivots, and we find that the latter works significantly better. Finally, we demonstrate that neither word-nor character-BLEU correlate perfectly with human judgments, due to BLEU's sensitivity to length.