{"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/sarg-a-novel-semi-autoregressive-generator","title":"SARG: A Novel Semi Autoregressive Generator for Multi-turn Incomplete Utterance Restoration","arxiv_id":"2008.01474","date":"2020-08-04","proceeding":null,"authors":["Mengzuo Huang","Feng Li","Wuhe Zou","Weidong Zhang"],"abstract":"Dialogue systems in open domain have achieved great success due to the easily obtained single-turn corpus and the development of deep learning, but the multi-turn scenario is still a challenge because of the frequent coreference and information omission. In this paper, we investigate the incomplete utterance restoration which has brought general improvement over multi-turn dialogue systems in recent studies. Meanwhile, jointly inspired by the autoregression for text generation and the sequence labeling for text editing, we propose a novel semi autoregressive generator (SARG) with the high efficiency and flexibility. Moreover, experiments on two benchmarks show that our proposed model significantly outperforms the state-of-the-art models in terms of quality and inference speed.","url_abs":"https://arxiv.org/abs/2008.01474v3","url_pdf":"https://arxiv.org/pdf/2008.01474v3.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":"sarg-a-novel-semi-autoregressive-generator","repo_url":"https://github.com/NetEase-GameAI/SARG","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"dialogue-rewriting","task_name":"Dialogue Rewriting"},{"task_slug":"text-generation","task_name":"Text Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/dialogue-rewriting-on-canard","task":"Dialogue Rewriting","dataset":"CANARD","model":"SARG","rank_in_archive_order":1,"of":1,"metrics":{"BLEU":"54.80"},"uses_additional_data":false},{"leaderboard":"/sota/dialogue-rewriting-on-multi-rewrite","task":"Dialogue Rewriting","dataset":"Multi-Rewrite","model":"SARG (n_beam=5)","rank_in_archive_order":2,"of":3,"metrics":{"Rewriting F2":"52.5","Rewriting F3":"46.4"},"uses_additional_data":false},{"leaderboard":"/sota/dialogue-rewriting-on-multi-rewrite","task":"Dialogue Rewriting","dataset":"Multi-Rewrite","model":"SARG (greedy)","rank_in_archive_order":3,"of":3,"metrics":{"BLEU-1":"92.2","BLEU-2":"89.6","ROUGE-1":"92.1","ROUGE-2":"86.0","Rewriting F1":"62.4"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2008.01474","atlas_url":"https://app.syntology.ai/?focus=2008.01474","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}