@inproceedings{briakou-etal-2022-bitextedit,
    title = "{B}itext{E}dit: Automatic Bitext Editing for Improved Low-Resource Machine Translation",
    author = "Briakou, Eleftheria  and
      Wang, Sida  and
      Zettlemoyer, Luke  and
      Ghazvininejad, Marjan",
    editor = "Carpuat, Marine  and
      de Marneffe, Marie-Catherine  and
      Meza Ruiz, Ivan Vladimir",
    booktitle = "Findings of the Association for Computational Linguistics: NAACL 2022",
    month = jul,
    year = "2022",
    address = "Seattle, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2022.findings-naacl.110/",
    doi = "10.18653/v1/2022.findings-naacl.110",
    pages = "1469--1485",
    abstract = "Mined bitexts can contain imperfect translations that yield unreliable training signals for Neural Machine Translation (NMT). While filtering such pairs out is known to improve final model quality, we argue that it is suboptimal in low-resource conditions where even mined data can be limited. In our work, we propose instead, to refine the mined bitexts via automatic editing: given a sentence in a language $x_f$, and a possibly imperfect translation of it $\mathbf{x_e}$, our model generates a revised version $x_f'$ or $x_e'$ that yields a more equivalent translation pair (i.e., {\ensuremath{<}}$x_f, x_e'${\ensuremath{>}} or {\ensuremath{<}}$x_f', x_e${\ensuremath{>}}). We use a simple editing strategy by (1) mining potentially imperfect translations for each sentence in a given bitext, (2) learning a model to reconstruct the original translations and translate, in a multi-task fashion. Experiments demonstrate that our approach successfully improves the quality of CCMatrix mined bitext for 5 low-resource language-pairs and 10 translation directions by up to 8 BLEU points, in most cases improving upon a competitive translation-based baseline."
}