{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T14:16:05Z","timestamp":1784556965655,"version":"3.55.0"},"reference-count":55,"publisher":"Association for Computing Machinery (ACM)","issue":"6","license":[{"start":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T00:00:00Z","timestamp":1784505600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["62406162 and 62506218"],"award-info":[{"award-number":["62406162 and 62506218"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"crossref","award":["2026MS031"],"award-info":[{"award-number":["2026MS031"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Fundamental Research Program of Shanxi Province","award":["202403021211092"],"award-info":[{"award-number":["202403021211092"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Inf. Syst."],"published-print":{"date-parts":[[2026,7,31]]},"abstract":"<jats:p>Event Relation Extraction (ERE) plays a crucial role in understanding document structures by identifying relations between events. However, most existing methods either rely on single-model instances, which often suffer from overconfidence, or adopt multi-agent frameworks that rely on manually designed prompts and heuristics to define agents, making effective optimization difficult. In this article, we propose Debate to Extract (D2E), a novel multi-agent optimization framework for ERE that leverages structured multi-turn debates and specialized agent training to enhance performance. Specifically, to organize the debate, participants are divided into multiple groups, each assigned its own debate topic. This process effectively integrates both cooperation and confrontation. We also incorporate an audience as a crucial participant, whose conclusions, from an observer\u2019s perspective, tend to be more objective. Building on this debate framework, D2E further optimizes the debate participants, combining structured multi-turn debates with agent training. During the debate, agents refine their initial opinions through collaborative interactions. This iterative process generates valuable supervision signals for training, allowing agents to improve their responses progressively. To address the issue of diminishing returns from data diversity, each agent is trained on distinct subsets of generated data, promoting specialization across different task dimensions. Experimental results on the MAVEN-ERE and EventStoryLine datasets show that D2E achieves significant improvements in causal relation extraction, outperforming baseline methods by 12.1% and 4.69%, respectively. Through further analysis of the debate participants\u2019 performance before and after the debate, we find that participation in the debate generally leads to improved ERE performance. This work demonstrates that combining collaborative debate with agent specialization leads to substantial performance gains in event relation extraction tasks.<\/jats:p>","DOI":"10.1145\/3819825","type":"journal-article","created":{"date-parts":[[2026,6,6]],"date-time":"2026-06-06T12:21:42Z","timestamp":1780748502000},"page":"1-28","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Debate to Extract: Enhancing Event Relation Extraction through Collaborative Debate and Agent Optimization"],"prefix":"10.1145","volume":"44","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9044-2595","authenticated-orcid":false,"given":"Yong","family":"Guan","sequence":"first","affiliation":[{"name":"North China Electric Power University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4130-3924","authenticated-orcid":false,"given":"Shaoru","family":"Guo","sequence":"additional","affiliation":[{"name":"Shanxi University, Shanxi, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8907-3526","authenticated-orcid":false,"given":"Lei","family":"Hou","sequence":"additional","affiliation":[{"name":"Tsinghua University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6244-0664","authenticated-orcid":false,"given":"Juanzi","family":"Li","sequence":"additional","affiliation":[{"name":"Tsinghua University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,7,20]]},"reference":[{"key":"e_1_3_2_2_2","volume-title":"Proceedings of the 2nd Workshop on Computing News Storylines","author":"Caselli Tommaso","year":"2016","unstructured":"Tommaso Caselli and Piek Vossen. 2016. The Storyline Annotation and Representation Scheme (StaR): A proposal. In Proceedings of the 2nd Workshop on Computing News Storylines. Retrieved from https:\/\/github.com\/tommasoc80\/EventStoryLine"},{"key":"e_1_3_2_3_2","doi-asserted-by":"crossref","first-page":"11534","DOI":"10.18653\/v1\/2025.findings-acl.601","volume-title":"Findings of the Association for Computational Linguistics (ACL \u201925)","author":"Chen Weize","year":"2025","unstructured":"Weize Chen, Jiarui Yuan, Chen Qian, Cheng Yang, Zhiyuan Liu, and Maosong Sun. 2025. Optima: Optimizing effectiveness and efficiency for LLM-based multi-agent system. In Findings of the Association for Computational Linguistics (ACL \u201925), 11534\u201311557. Retrieved from https:\/\/aclanthology.org\/2025.findings-acl.601\/"},{"key":"e_1_3_2_4_2","unstructured":"Josef Dai Xuehai Pan Ruiyang Sun Jiaming Ji Xinbo Xu Mickel Liu Yizhou Wang and Yaodong Yang. 2023. Safe RLHF: Safe reinforcement learning from human feedback. arXiv:2310.12773. Retrieved from https:\/\/arxiv.org\/abs\/2310.12773"},{"key":"e_1_3_2_5_2","unstructured":"Heshan Fernando Han Shen Parikshit Ram Yi Zhou Horst Samulowitz Nathalie Baracaldo and Tianyi Chen. 2024. Mitigating forgetting in LLM supervised fine-tuning and preference learning. arXiv:2410.15483. Retrieved from https:\/\/arxiv.org\/abs\/2410.15483"},{"key":"e_1_3_2_6_2","unstructured":"Aaron Grattafiori Abhimanyu Dubey Abhinav Jauhri Abhinav Pandey Abhishek Kadian Ahmad Al-Dahle Aiesha Letman Akhil Mathur Alan Schelten Alex Vaughan Amy Yang et al. 2024. The Llama 3 herd of models. arXiv:2407.21783. Retrieved from https:\/\/arxiv.org\/abs\/2407.21783"},{"key":"e_1_3_2_7_2","first-page":"6889","volume-title":"Proceedings of the 31st International Conference on Computational Linguistics","author":"Guan Yong","year":"2025","unstructured":"Yong Guan, Hao Peng, Lei Hou, and Juanzi Li. 2025. MMD-ERE: Multi-agent multi-sided debate for event relation extraction. In Proceedings of the 31st International Conference on Computational Linguistics, 6889\u20136896. Retrieved from https:\/\/aclanthology.org\/2025.coling-main.460\/"},{"key":"e_1_3_2_8_2","first-page":"15511","volume-title":"Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation","author":"Guan Yong","year":"2024","unstructured":"Yong Guan, Xiaozhi Wang, Lei Hou, Juanzi Li, Jeff Z. Pan, Jiaoyan Chen, and Freddy Lecue. 2024. TacoERE: Cluster-aware compression for event relation extraction. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, 15511\u201315521. Retrieved from https:\/\/aclanthology.org\/2024.lrec-main.1348\/"},{"key":"e_1_3_2_9_2","unstructured":"Ridong Han Tao Peng Chaohao Yang Benyou Wang Lu Liu and Xiang Wan. 2023. Is information extraction solved by ChatGPT? An analysis of performance evaluation criteria robustness and errors. arXiv:2305.14450. Retrieved from https:\/\/arxiv.org\/abs\/2305.14450"},{"key":"e_1_3_2_10_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-acl.507"},{"key":"e_1_3_2_11_2","unstructured":"Edward J. Hu Yelong Shen Phillip Wallis Zeyuan Allen-Zhu Yuanzhi Li Shean Wang Lu Wang and Weizhu Chen. 2021. LoRA: Low-rank adaptation of large language models. arXiv:2106.09685. Retrieved from https:\/\/arxiv.org\/abs\/2106.09685"},{"key":"e_1_3_2_12_2","first-page":"7484","volume-title":"Proceedings of the 31st International Conference on Computational Linguistics","author":"Hu Zhilei","year":"2025","unstructured":"Zhilei Hu, Zixuan Li, Xiaolong Jin, Long Bai, Jiafeng Guo, and Xueqi Cheng. 2025. Large language model-based event relation extraction with rationales. In Proceedings of the 31st International Conference on Computational Linguistics, 7484\u20137496. Retrieved from https:\/\/aclanthology.org\/2025.coling-main.500\/"},{"key":"e_1_3_2_13_2","doi-asserted-by":"publisher","DOI":"10.1109\/TVT.2022.3165526"},{"key":"e_1_3_2_14_2","unstructured":"Haochen Li Wanjin Feng Xin Zhou and Zhiqi Shen. 2025. GiFT: Gibbs fine-tuning for code generation. arXiv:2502.11466. Retrieved from https:\/\/arxiv.org\/abs\/2502.11466"},{"key":"e_1_3_2_15_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.1105"},{"issue":"7","key":"e_1_3_2_16_2","first-page":"7118","article-title":"Learning generative RNN-ODE for collaborative time-series and event sequence forecasting","volume":"35","author":"Li Longyuan","year":"2022","unstructured":"Longyuan Li, Junchi Yan, Yunhao Zhang, Jihai Zhang, Jie Bao, Yaohui Jin, and Xiaokang Yang. 2022. Learning generative RNN-ODE for collaborative time-series and event sequence forecasting. IEEE Transactions on Knowledge and Data Engineering 35, 7 (2022), 7118\u20137137.","journal-title":"IEEE Transactions on Knowledge and Data Engineering"},{"key":"e_1_3_2_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/3583780.3615017"},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.992"},{"key":"e_1_3_2_19_2","first-page":"7201","volume-title":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics","author":"Lu Yi-Fan","year":"2025","unstructured":"Yi-Fan Lu, Xian-Ling Mao, Tian Lan, Tong Zhang, Yu-Shi Zhu, and He-Yan Huang. 2025. SEOE: A scalable and reliable semantic evaluation framework for open domain event detection. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics, 7201\u20137218."},{"key":"e_1_3_2_20_2","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3403373"},{"key":"e_1_3_2_21_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.procs.2015.09.168"},{"key":"e_1_3_2_22_2","unstructured":"OpenAI. 2023. GPT-4 technical report. arXiv:2303.08774. Retrieved from https:\/\/arxiv.org\/abs\/2303.08774"},{"key":"e_1_3_2_23_2","unstructured":"Jinlong Pang Na Di Zhaowei Zhu Jiaheng Wei Hao Cheng Chen Qian and Yang Liu. 2025. Token cleaning: Fine-grained data selection for LLM supervised fine-tuning. arXiv:2502.01968. Retrieved from https:\/\/arxiv.org\/abs\/2502.01968"},{"key":"e_1_3_2_24_2","unstructured":"Hao Peng Xiaozhi Wang Jianhui Chen Weikai Li Yunjia Qi Zimu Wang Zhili Wu Kaisheng Zeng Bin Xu Lei Hou and Juanzi Li. 2023. When does in-context learning fall short and why? A study on specification-heavy tasks. arXiv:2311.08993. Retrieved from https:\/\/arxiv.org\/abs\/2311.08993"},{"key":"e_1_3_2_25_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.419"},{"key":"e_1_3_2_26_2","volume-title":"Advances in Neural Information Processing Systems","volume":"36","author":"Rafailov Rafael","year":"2023","unstructured":"Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D. Manning, Stefano Ermon, and Chelsea Finn. 2023. Direct preference optimization: Your language model is secretly a reward model. In Advances in Neural Information Processing Systems, Vol. 36. Retrieved from https:\/\/papers.nips.cc\/paper_files\/paper\/2023\/hash\/a85b 405ed65c6477a4fe8302b5e06ce7-Abstract-Conference.html"},{"key":"e_1_3_2_27_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW56347.2022.00301"},{"key":"e_1_3_2_28_2","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0377"},{"key":"e_1_3_2_29_2","unstructured":"Yuda Song Hanlin Zhang Carson Eisenach Sham Kakade Dean Foster and Udaya Ghai. 2025. Mind the gap: Examining the self-improvement capabilities of large language models. arXiv:2412.02674. Retrieved from https:\/\/arxiv.org\/abs\/2412.02674"},{"key":"e_1_3_2_30_2","unstructured":"Yuda Song Hanlin Zhang Carson Eisenach Sham M. Kakade Dean P. Foster and Udaya Ghai. 2024. Mind the gap: Examining the self-improvement capabilities of large language models. arXiv:2412.02674. Retrieved from https:\/\/doi.org\/10.48550\/ARXIV.2412.02674"},{"key":"e_1_3_2_31_2","first-page":"32764","volume-title":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing","author":"Srivastava Gaurav","year":"2025","unstructured":"Gaurav Srivastava, Zhenyu Bi, Meng Lu, and Xuan Wang. 2025. DEBATE, TRAIN, EVOLVE: Self-evolution of language model reasoning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 32764\u201332810. Retrieved from https:\/\/aclanthology.org\/2025.emnlp-main.1666\/"},{"key":"e_1_3_2_32_2","doi-asserted-by":"publisher","DOI":"10.1007\/s10626-025-00421-w"},{"key":"e_1_3_2_33_2","unstructured":"Aohan Zeng Xin Lv Qinkai Zheng Zhenyu Hou Bin Chen Chengxing Xie Cunxiang Wang Da Yin Hao Zeng Jiajie Zhang et al. 2025. GLM-4.5: Agentic Reasoning and Coding (ARC) foundation models. arXiv:2508.06471. Retrieved from https:\/\/arxiv.org\/abs\/2508.06471"},{"key":"e_1_3_2_34_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.naacl-main.273"},{"key":"e_1_3_2_35_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.1335"},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.emnlp-main.60"},{"key":"e_1_3_2_37_2","doi-asserted-by":"crossref","unstructured":"Xiao Wang Jingtao Jiang Qiang Chen Lan Chen Lin Zhu Yaowei Wang Yonghong Tian and Jin Tang. 2025. ESTR-CoT: Towards explainable and accurate event stream based scene text recognition with chain-of-thought reasoning. arXiv:2507.02200. Retrieved from https:\/\/arxiv.org\/abs\/2507.02200","DOI":"10.2139\/ssrn.5770753"},{"key":"e_1_3_2_38_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.224"},{"key":"e_1_3_2_39_2","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-023-00691-9"},{"key":"e_1_3_2_40_2","doi-asserted-by":"publisher","DOI":"10.3390\/bdcc9040087"},{"key":"e_1_3_2_41_2","unstructured":"Zhiheng Xi Wenxiang Chen Xin Guo Wei He Yiwen Ding Boyang Hong Ming Zhang Junzhe Wang Senjie Jin Enyu Zhou et al. 2023. The rise and potential of large language model based agents: A survey. arXiv:2309.07864. Retrieved from https:\/\/arxiv.org\/abs\/2309.07864"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","DOI":"10.3390\/electronics12153349"},{"key":"e_1_3_2_43_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298789"},{"key":"e_1_3_2_44_2","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Yao Shunyu","year":"2022","unstructured":"Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik R. Narasimhan, and Yuan Cao. 2022. React: Synergizing reasoning and acting in language models. In Proceedings of the International Conference on Learning Representations."},{"key":"e_1_3_2_45_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.936"},{"key":"e_1_3_2_46_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.naacl-long.287"},{"key":"e_1_3_2_47_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.acl-long.897"},{"key":"e_1_3_2_48_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.bionlp-1.7"},{"key":"e_1_3_2_49_2","volume-title":"Proceedings of the 36th Conference on Neural Information Processing Systems","author":"Zelikman Eric","year":"2024","unstructured":"Eric Zelikman, Yuhuai Wu, Jesse Mu, and Noah D. Goodman. 2024. Star: Self-taught reasoner bootstrapping reasoning with reasoning. In Proceedings of the 36th Conference on Neural Information Processing Systems, Article 1126."},{"key":"e_1_3_2_50_2","first-page":"16810","volume-title":"Findings of the Association for Computational Linguistics","author":"Zhang Shaowei","year":"2025","unstructured":"Shaowei Zhang and Deyi Xiong. 2025. Debate4MATH: Multi-agent debate for fine-grained reasoning in math. In Findings of the Association for Computational Linguistics, 16810\u201316824. Retrieved from https:\/\/aclanthology.org\/2025.findings-acl.862\/"},{"key":"e_1_3_2_51_2","unstructured":"Xinliang Frederick Zhang Carter Blum Temma Choji Shalin Shah and Alakananda Vempala. 2024. ULTRA: Unleash LLMs\u2019 potential for event argument extraction through hierarchical modeling and pair-wise refinement. arXiv:2401.13218. Retrieved from https:\/\/arxiv.org\/abs\/2401.13218"},{"key":"e_1_3_2_52_2","unstructured":"Wanjia Zhao Mert Yuksekgonul Shirley Wu and James Zou. 2025. SiriuS: Self-improving multi-agent systems via bootstrapped reasoning. arXiv:2502.04780. Retrieved from https:\/\/arxiv.org\/abs\/2502.04780"},{"key":"e_1_3_2_53_2","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2024\/736"},{"key":"e_1_3_2_54_2","unstructured":"Han Zhong Zikang Shan Guhao Feng Wei Xiong Xinle Cheng Li Zhao Di He Jiang Bian and Liwei Wang. 2024. DPO meets PPO: Reinforced token optimization FOR RLHF. arXiv:2404.18922. Retrieved from https:\/\/arxiv.org\/abs\/2404.18922"},{"key":"e_1_3_2_55_2","first-page":"9122","volume-title":"Findings of the ACL","author":"Zhou Xiaofeng","year":"2025","unstructured":"Xiaofeng Zhou, Heyan Huang, and Lizi Liao. 2025. Debate, reflect, and distill: Multi-agent feedback with tree-structured preference optimization for efficient language model enhancement. In Findings of the ACL, 9122\u20139137. Retrieved from https:\/\/aclanthology.org\/2025.findings-acl.475\/"},{"issue":"6","key":"e_1_3_2_56_2","doi-asserted-by":"crossref","first-page":"3461","DOI":"10.1109\/TSMC.2022.3225381","article-title":"An optimal iterative learning control approach for linear systems with nonuniform trial lengths under input constraints","volume":"53","author":"Zhuang Zhihe","year":"2022","unstructured":"Zhihe Zhuang, Hongfeng Tao, Yiyang Chen, Vladimir Stojanovic, and Wojciech Paszke. 2022. An optimal iterative learning control approach for linear systems with nonuniform trial lengths under input constraints. IEEE Transactions on Systems, Man, and Cybernetics: Systems 53, 6 (2022), 3461\u20133473.","journal-title":"IEEE Transactions on Systems, Man, and Cybernetics: Systems"}],"container-title":["ACM Transactions on Information Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3819825","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T13:24:58Z","timestamp":1784553898000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3819825"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,20]]},"references-count":55,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2026,7,31]]}},"alternative-id":["10.1145\/3819825"],"URL":"https:\/\/doi.org\/10.1145\/3819825","relation":{},"ISSN":["1046-8188","1558-2868"],"issn-type":[{"value":"1046-8188","type":"print"},{"value":"1558-2868","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,20]]},"assertion":[{"value":"2025-11-25","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-05-13","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-07-20","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}