{
  "id": "2005.04816",
  "version": "v1",
  "published": "2020-05-11T00:20:33.000Z",
  "updated": "2020-05-11T00:20:33.000Z",
  "title": "Leveraging Monolingual Data with Self-Supervision for Multilingual Neural Machine Translation",
  "authors": [
    "Aditya Siddhant",
    "Ankur Bapna",
    "Yuan Cao",
    "Orhan Firat",
    "Mia Chen",
    "Sneha Kudugunta",
    "Naveen Arivazhagan",
    "Yonghui Wu"
  ],
  "journal": "ACL 2020",
  "categories": [
    "cs.CL",
    "cs.LG"
  ],
  "abstract": "Over the last few years two promising research directions in low-resource neural machine translation (NMT) have emerged. The first focuses on utilizing high-resource languages to improve the quality of low-resource languages via multilingual NMT. The second direction employs monolingual data with self-supervision to pre-train translation models, followed by fine-tuning on small amounts of supervised data. In this work, we join these two lines of research and demonstrate the efficacy of monolingual data with self-supervision in multilingual NMT. We offer three major results: (i) Using monolingual data significantly boosts the translation quality of low-resource languages in multilingual models. (ii) Self-supervision improves zero-shot translation quality in multilingual models. (iii) Leveraging monolingual data with self-supervision provides a viable path towards adding new languages to multilingual models, getting up to 33 BLEU on ro-en translation without any parallel data or back-translation.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2020-05-11T00:20:33.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "multilingual neural machine translation",
      "leveraging monolingual data",
      "self-supervision",
      "multilingual models",
      "second direction employs monolingual data"
    ],
    "tags": [
      "journal article"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}