{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/using-monolingual-data-in-neural-machine-1","title":"Using Monolingual Data in Neural Machine Translation: a Systematic Study","arxiv_id":"1903.11437","date":"2019-03-27","proceeding":"WS 2018 10","authors":["Franck Burlot","François Yvon"],"abstract":"Neural Machine Translation (MT) has radically changed the way systems are\ndeveloped. A major difference with the previous generation (Phrase-Based MT) is\nthe way monolingual target data, which often abounds, is used in these two\nparadigms. While Phrase-Based MT can seamlessly integrate very large language\nmodels trained on billions of sentences, the best option for Neural MT\ndevelopers seems to be the generation of artificial parallel data through\n\\textsl{back-translation} - a technique that fails to fully take advantage of\nexisting datasets. In this paper, we conduct a systematic study of\nback-translation, comparing alternative uses of monolingual data, as well as\nmultiple data generation procedures. Our findings confirm that back-translation\nis very effective and give new explanations as to why this is the case. We also\nintroduce new data simulation techniques that are almost as effective, yet much\ncheaper to implement.","url_abs":"http://arxiv.org/abs/1903.11437v1","url_pdf":"http://arxiv.org/pdf/1903.11437v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"using-monolingual-data-in-neural-machine-1","repo_url":"https://github.com/franckbrl/nmt-pseudo-source-discriminator","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.11437","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}