{"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/towards-neural-phrase-based-machine","title":"Towards Neural Phrase-based Machine Translation","arxiv_id":"1706.05565","date":"2017-06-17","proceeding":"ICLR 2018 1","authors":["Po-Sen Huang","Chong Wang","Sitao Huang","Dengyong Zhou","Li Deng"],"abstract":"In this paper, we present Neural Phrase-based Machine Translation (NPMT). Our\nmethod explicitly models the phrase structures in output sequences using\nSleep-WAke Networks (SWAN), a recently proposed segmentation-based sequence\nmodeling method. To mitigate the monotonic alignment requirement of SWAN, we\nintroduce a new layer to perform (soft) local reordering of input sequences.\nDifferent from existing neural machine translation (NMT) approaches, NPMT does\nnot use attention-based decoding mechanisms. Instead, it directly outputs\nphrases in a sequential order and can decode in linear time. Our experiments\nshow that NPMT achieves superior performances on IWSLT 2014\nGerman-English/English-German and IWSLT 2015 English-Vietnamese machine\ntranslation tasks compared with strong NMT baselines. We also observe that our\nmethod produces meaningful phrases in output languages.","url_abs":"http://arxiv.org/abs/1706.05565v8","url_pdf":"http://arxiv.org/pdf/1706.05565v8.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":"towards-neural-phrase-based-machine","repo_url":"https://github.com/posenhuang/NPMT","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"torch","reach":{"status":"unanswered"}},{"paper_slug":"towards-neural-phrase-based-machine","repo_url":"https://github.com/Microsoft/NPMT","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"torch","reach":{"status":"unanswered"}},{"paper_slug":"towards-neural-phrase-based-machine","repo_url":"https://github.com/ykrmm/ICLR_2020","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"towards-neural-phrase-based-machine","repo_url":"https://github.com/ykrmm/TREMBA","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"nmt","task_name":"NMT"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/machine-translation-on-iwslt2014-german","task":"Machine Translation","dataset":"IWSLT2014 German-English","model":"Neural PBMT + LM [Huang2018]","rank_in_archive_order":32,"of":34,"metrics":{"BLEU score":"30.08"},"uses_additional_data":false},{"leaderboard":"/sota/machine-translation-on-iwslt2015-english","task":"Machine Translation","dataset":"IWSLT2015 English-German","model":"NPMT + language model","rank_in_archive_order":7,"of":8,"metrics":{"BLEU score":"25.36"},"uses_additional_data":false},{"leaderboard":"/sota/machine-translation-on-iwslt2015-german","task":"Machine Translation","dataset":"IWSLT2015 German-English","model":"NPMT + language model","rank_in_archive_order":8,"of":15,"metrics":{"BLEU score":"30.08"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.05565","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}