{"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/near-human-level-performance-in-grammatical","title":"Near Human-Level Performance in Grammatical Error Correction with Hybrid Machine Translation","arxiv_id":"1804.05945","date":"2018-04-16","proceeding":"NAACL 2018 6","authors":["Roman Grundkiewicz","Marcin Junczys-Dowmunt"],"abstract":"We combine two of the most popular approaches to automated Grammatical Error\nCorrection (GEC): GEC based on Statistical Machine Translation (SMT) and GEC\nbased on Neural Machine Translation (NMT). The hybrid system achieves new\nstate-of-the-art results on the CoNLL-2014 and JFLEG benchmarks. This GEC\nsystem preserves the accuracy of SMT output and, at the same time, generates\nmore fluent sentences as it typical for NMT. Our analysis shows that the\ncreated systems are closer to reaching human-level performance than any other\nGEC system reported so far.","url_abs":"http://arxiv.org/abs/1804.05945v1","url_pdf":"http://arxiv.org/pdf/1804.05945v1.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":[],"tasks":[{"task_slug":"grammatical-error-correction","task_name":"Grammatical Error Correction"},{"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/grammatical-error-correction-on-conll-2014","task":"Grammatical Error Correction","dataset":"CoNLL-2014 Shared Task","model":"SMT + BiGRU","rank_in_archive_order":21,"of":23,"metrics":{"F0.5":"56.25"},"uses_additional_data":false},{"leaderboard":"/sota/grammatical-error-correction-on-conll-2014-1","task":"Grammatical Error Correction","dataset":"CoNLL-2014 Shared Task (10 annotations)","model":"SMT + BiGRU","rank_in_archive_order":2,"of":3,"metrics":{"F0.5":"72.04 "},"uses_additional_data":false},{"leaderboard":"/sota/grammatical-error-correction-on-jfleg","task":"Grammatical Error Correction","dataset":"JFLEG","model":"SMT + BiGRU","rank_in_archive_order":3,"of":6,"metrics":{"GLEU":"61.5"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1804.05945","atlas_url":"https://app.syntology.ai/?focus=1804.05945","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}