@inproceedings{zouhar-etal-2024-pitfalls,
    title = "Pitfalls and Outlooks in Using {COMET}",
    author = "Zouhar, Vil{\'e}m  and
      Chen, Pinzhen  and
      Lam, Tsz Kin  and
      Moghe, Nikita  and
      Haddow, Barry",
    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-1.121/",
    doi = "10.18653/v1/2024.wmt-1.121",
    pages = "1272--1288",
    abstract = "The COMET metric has blazed a trail in the machine translation community, given its strong correlation with human judgements of translation quality.Its success stems from being a modified pre-trained multilingual model finetuned for quality assessment.However, it being a machine learning model also gives rise to a new set of pitfalls that may not be widely known. We investigate these unexpected behaviours from three aspects:1) technical: obsolete software versions and compute precision; 2) data: empty content, language mismatch, and translationese at test time as well as distribution and domain biases in training; 3) usage and reporting: multi-reference support and model referencing in the literature. All of these problems imply that COMET scores are not comparable between papers or even technical setups and we put forward our perspective on fixing each issue.Furthermore, we release the sacreCOMET package that can generate a signature for the software and model configuration as well as an appropriate citation.The goal of this work is to help the community make more sound use of the COMET metric."
}