@inproceedings{post-junczys-dowmunt-2024-evaluation,
    title = "Evaluation and Large-scale Training for Contextual Machine Translation",
    author = "Post, Matt  and
      Junczys-Dowmunt, Marcin",
    editor = "Haddow, Barry  and
      Kocmi, Tom  and
      Koehn, Philipp  and
      Monz, Christof",
    booktitle = "Proceedings of the Ninth Conference on Machine Translation",
    month = nov,
    year = "2024",
    address = "Miami, Florida, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2024.wmt-1.112/",
    doi = "10.18653/v1/2024.wmt-1.112",
    pages = "1125--1139",
    abstract = "Despite the fact that context is known to be vital for resolving a range of translation ambiguities, most traditional machine translation systems continue to be trained and to operate at the sentence level. A common explanation is the lack of document-level annotations for existing training data. This work investigates whether having such annotations would be helpful for training traditional MT systems at scale. We build large-scale, state-of-the-art contextual MT systems into German, French, and Russian, fixing the datasets while comparing the effect of sourcing contextual training samples from both parallel and back-translated data. We then evaluate these contextual models across a range of contextual test sets from the literature, where we find that (a) document annotations from both mined parallel and back-translated monolingual data are helpful, but that the best contextual MT systems do not draw contextual samples from the parallel data. We also make two points related to evaluation: (b) contrastive score-based metrics on challenge sets are not discriminative; instead, models must be tested directly on their ability to generate correct outputs, and (c) standard corpus-level metrics such as COMET work best in settings that are dense in contextual phenomena."
}