@inproceedings{koretaka-etal-2023-mitigating,
    title = "Mitigating Domain Mismatch in Machine Translation via Paraphrasing",
    author = "Koretaka, Hyuga  and
      Kajiwara, Tomoyuki  and
      Fujita, Atsushi  and
      Ninomiya, Takashi",
    editor = "Nakazawa, Toshiaki  and
      Kinugawa, Kazutaka  and
      Mino, Hideya  and
      Goto, Isao  and
      Dabre, Raj  and
      Higashiyama, Shohei  and
      Parida, Shantipriya  and
      Morishita, Makoto  and
      Bojar, Ondrej  and
      Eriguchi, Akiko  and
      Oda, Yusuke  and
      Eriguchi, Akiko  and
      Chu, Chenhui  and
      Kurohashi, Sadao",
    booktitle = "Proceedings of the 10th Workshop on Asian Translation",
    month = sep,
    year = "2023",
    address = "Macau SAR, China",
    publisher = "Asia-Pacific Association for Machine Translation",
    url = "https://aclanthology.org/2023.wat-1.2",
    pages = "29--40",
    abstract = "Quality of machine translation (MT) deteriorates significantly when translating texts having characteristics that differ from the training data, such as content domain. Although previous studies have focused on adapting MT models on a bilingual parallel corpus in the target domain, this approach is not applicable when no parallel data are available for the target domain or when utilizing black-box MT systems. To mitigate problems caused by such domain mismatch without relying on any corpus in the target domain, this study proposes a method to search for better translations by paraphrasing input texts of MT. To obtain better translations even for input texts from unforeknown domains, we generate their multiple paraphrases, translate each, and rerank the resulting translations to select the most likely one. Experimental results on Japanese-to-English translation reveal that the proposed method improves translation quality in terms of BLEU score for input texts from specific domains.",
}