---
title: "Weakly-supervised Argument Mining with Boundary Refinement"
description: "Argument mining (AM) turns paragraph text into argumentative graphs. 25 references. Among them: 0."
authors: ["Wei Sun", "Mingxiao Li", "Jesse Davis", "Elena Cabrio", "Serena Villata", "Marie-Francine Moens"]
year: 2026
venue: "Proceedings of the Ninth Fact Extraction and VERification Workshop (FEVER)"
source: "https://aclanthology.org/2026.fever-1.1/"
doi: "10.18653/v1/2026.fever-1.1"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/sun-2026-weakly-supervised-argument-mining-boundary/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/sun-2026-weakly-supervised-argument-mining-boundary/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/47/) / Weakly-supervised Argument Mining with Boundary Refinement and Relation Denoising

# Weakly-supervised Argument Mining with Boundary Refinement and Relation Denoising

**Authors:** [Wei Sun](https://md.research.elamur.ai/free/en/authors/wei-sun/), Mingxiao Li, Jesse Davis, [Elena Cabrio](https://md.research.elamur.ai/free/en/authors/elena-cabrio/), [Serena Villata](https://md.research.elamur.ai/free/en/authors/serena-villata/) and [Marie-Francine Moens](https://md.research.elamur.ai/free/en/authors/marie-francine-moens/) · **Year:** 2026 · **Venue:** [Proceedings of the Ninth Fact Extraction and VERification Workshop (FEVER)](https://md.research.elamur.ai/free/en/venues/fever/) ([FEVER 2026](https://md.research.elamur.ai/free/en/venues/fever/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.fever-1.1/) · [PDF](https://aclanthology.org/2026.fever-1.1.pdf) · [DOI 10.18653/v1/2026.fever-1.1](https://doi.org/10.18653/v1/2026.fever-1.1) · [OpenAlex](https://openalex.org/W7140074205)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Argument mining (AM) turns paragraph text into argumentative graphs. Spans are labelled as components such as claims or premises and then linked by relations like support or attack. Strong AM systems usually depend on large amounts of manual annotation. The authors instead have an LLM pseudo-label the training data through few-shot in-context prompting, and they add two modules to cope with the resulting noise.

The first is a center-based component detector built on SciBERT. Rather than predicting hard BIO tags, which often get component starts wrong, it learns soft Gaussian-shaped masks marking each span's center and edges. The second is a relation denoiser. It trains the relation classifier on two weak label sources, the LLM's annotations and those of a model tuned on the gold development split, and it reweights the two sources dynamically according to how much they agree.

Across four public datasets, the approach sets a new state of the art for weakly-supervised AM. It reaches fully-supervised quality using less than a fifth of the human-labelled data, which makes graph extraction cheaper in costly domains like clinical text.

## Key points
- Evaluation covers four public benchmarks: AbstRCT-neoplasm, AbstRCT-glaucoma, AbstRCT-mixed and CDCP.
- Compared with the raw LLM pseudo-labels, boundary detection gains up to 16% IoU 75 and relation classification up to 12% macro-F1.
- Against the previous best weakly-supervised systems, the model improves component detection by up to 6% and relation classification by up to 7%. With under 20% of the gold annotations it equals fully-supervised AM, including the Mayer et al. (2020) pipeline baseline.
- Sub-models are trained with Continuous Fine-Tuning (CFT), first on the weakly labelled data and then on the gold development set. Center and boundary masks use a masked MSE loss restricted to each word's first sub-token, and component labels use a cross-entropy loss.
- The stated limitations are that the work covers only paragraph-level input, since LLM labels for whole documents may be too noisy, and that it uses only plain few-shot prompting. CoT, PS-CoT and ToT prompting are left for future work.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Institut national de recherche en sciences et technologies du numérique](https://md.research.elamur.ai/free/en/labs/institut-national-de-recherche-en-sciences-et-technologies-du-numerique/) — Research institute, Le Chesnay-Rocquencourt, France · 84 works · 25,414 citations
- [Centre National de la Recherche Scientifique](https://md.research.elamur.ai/free/en/labs/centre-national-de-la-recherche-scientifique/) — Research institute, Paris, France · 168 works · 62,369 citations
- [KU Leuven](https://md.research.elamur.ai/free/en/universities/ku-leuven/) — University, Leuven, Belgium · 42 works · 5,839 citations
- [Laboratoire d'Informatique, Signaux et Systèmes de Sophia Antipolis](https://md.research.elamur.ai/free/en/labs/laboratoire-d-informatique-signaux-et-systemes-de-sophia-antipolis/) — Research institute, Valbonne, France · 3 works · 1 citation
- [Université Côte d'Azur](https://md.research.elamur.ai/free/en/universities/universite-cote-d-azur/) — University, Nice, France · 3 works · 1 citation

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [🤔 Philosophy](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/philosophy/) · 284 works

## Cite

Cite as: Wei Sun, Mingxiao Li, Jesse Davis, Elena Cabrio, Serena Villata and Marie-Francine Moens. 2026. Weakly-supervised Argument Mining with Boundary Refinement and Relation Denoising. Proceedings of the Ninth Fact Extraction and VERification Workshop (FEVER).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{sun2026weakly,
  title = {{Weakly-supervised Argument Mining with Boundary Refinement and Relation Denoising}},
  author = {Wei Sun, Mingxiao Li, Jesse Davis, Elena Cabrio, Serena Villata and Marie-Francine Moens},
  booktitle = {Proceedings of the Ninth Fact Extraction and VERification Workshop (FEVER)},
  year = {2026},
  doi = {10.18653/v1/2026.fever-1.1},
  url = {https://doi.org/10.18653/v1/2026.fever-1.1}
}
```

- [← previous We Are What We Repeatedly Do: Improving Long Context Instruction Following](https://md.research.elamur.ai/free/en/papers/robinette-2026-we-are-what-we-repeatedly/)
- [next → WebClipper: Efficient Evolution of Web Agents with Graph-based Trajectory Pruning](https://md.research.elamur.ai/free/en/papers/wang-2026-webclipper-efficient-evolution-web-agents/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/47/)

## References

25 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Jianzhu Bao, Chuang Fan, Jipeng Wu, Yixue Dang, Jiachen Du, and Ruifeng Xu. 2021. A neural transition-based model for argumentation mining. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Proce… [find](https://scholar.google.com/scholar?q=Jianzhu+Bao%2C+Chuang+Fan%2C+Jipeng+Wu%2C+Yixue+Dang%2C+Jiachen+Du%2C+and+Ruifeng+Xu%2E+2021%2E+A+neural+transition%2Dbased+model+for+argumentation+mining%2E+In+Proceedings+of+the+59th+Annual+Meeting+of+the+Association)
2. Wei Liu, Irtiza Hasan, and Shengcai Liao. 2020. Cen-terand scale prediction: A box-free approach for pedestrianand face detection. In Computer Vision and Pattern Recognition. [find](https://scholar.google.com/scholar?q=Wei+Liu%2C+Irtiza+Hasan%2C+and+Shengcai+Liao%2E+2020%2E+Cen%2Dterand+scale+prediction%3A+A+box%2Dfree+approach+for+pedestrianand+face+detection%2E+In+Computer+Vision+and+Pattern+Recognition%2E)
3. Iz Beltagy, Kyle Lo, and Arman Cohan. 2019. SciB-ERT: A pretrained language model for scientific text. In Proceedings of the 2019 Conference on Empirical Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Iz+Beltagy%2C+Kyle+Lo%2C+and+Arman+Cohan%2E+2019%2E+SciB%2DERT%3A+A+pretrained+language+model+for+scientific+text%2E+In+Proceedings+of+the+2019+Conference+on+Empirical+Ilya+Loshchilov+and+Frank+Hutter%2E+2019%2E+Decoup)
4. Yaojie Lu, Qing Liu, Dai Dai, Xinyan Xiao, Hongyu Lin, Xianpei Han, Le Sun, and Hua Wu. 2022. Unified structure generation for universal information extraction. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5755–5772, Dublin… [find](https://scholar.google.com/scholar?q=Yaojie+Lu%2C+Qing+Liu%2C+Dai+Dai%2C+Xinyan+Xiao%2C+Hongyu+Lin%2C+Xianpei+Han%2C+Le+Sun%2C+and+Hua+Wu%2E+2022%2E+Unified+structure+generation+for+universal+information+extraction%2E+In+Proceedings+of+the+60th+Annual+Meeti)
5. Lei Wang, Wanyu Xu, Yihuai Lan, Zhiqiang Hu, Yunshi Lan, Roy Ka-Wei Lee, and Ee-Peng Lim. 2023a. Plan-and-solve prompting: Improving zero-shot chain-of-thought reasoning by large language models. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Lo… [find](https://scholar.google.com/scholar?q=Lei+Wang%2C+Wanyu+Xu%2C+Yihuai+Lan%2C+Zhiqiang+Hu%2C+Yunshi+Lan%2C+Roy+Ka%2DWei+Lee%2C+and+Ee%2DPeng+Lim%2E+2023a%2E+Plan%2Dand%2Dsolve+prompting%3A+Improving+zero%2Dshot+chain%2Dof%2Dthought+reasoning+by+large+language+models%2E+In+P)
6. Tobias Mayer, Elena Cabrio, and Serena Villata. 2020. Transformer-based argument mining for healthcare applications. In ECAI 2020, pages 2108–2115. IOS Press. [find](https://scholar.google.com/scholar?q=Tobias+Mayer%2C+Elena+Cabrio%2C+and+Serena+Villata%2E+2020%2E+Transformer%2Dbased+argument+mining+for+healthcare+applications%2E+In+ECAI+2020%2C+pages+2108%E2%80%932115%2E+IOS+Press%2E)
7. Risheng Wang, Tao Lei, Ruixia Cui, Bingtao Zhang, Hongying Meng, and Asoke K Nandi. 2022. Medical image segmentation using deep learning: A survey. IET Image Processing, 16(5):1243–1267. [find](https://scholar.google.com/scholar?q=Risheng+Wang%2C+Tao+Lei%2C+Ruixia+Cui%2C+Bingtao+Zhang%2C+Hongying+Meng%2C+and+Asoke+K+Nandi%2E+2022%2E+Medical+image+segmentation+using+deep+learning%3A+A+survey%2E+IET+Image+Processing%2C+16%285%29%3A1243%E2%80%931267%2E)
8. Tristan Miller, Maria Sukhareva, and Iryna Gurevych. 2019. A streamlined method for sourcing discourse-level argumentation annotations from the crowd. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Vo… [find](https://scholar.google.com/scholar?q=Tristan+Miller%2C+Maria+Sukhareva%2C+and+Iryna+Gurevych%2E+2019%2E+A+streamlined+method+for+sourcing+discourse%2Dlevel+argumentation+annotations+from+the+crowd%2E+In+Proceedings+of+the+2019+Conference+of+the+Nort)
9. Shuhe Wang, Xiaofei Sun, Xiaoya Li, Rongbin Ouyang, Fei Wu, Tianwei Zhang, Jiwei Li, and Guoyin Wang. 2023b. [Gpt-ner: Named entity recognition via large language models](https://md.research.elamur.ai/free/en/papers/wang-2025-gpt-ner-named-entity-recognition/). arXiv preprint arXiv:2304.10428.
10. Yijun Wang, Changzhi Sun, Yuanbin Wu, Junchi Yan, Peng Gao, and Guotong Xie. 2020. Pre-training entity relation encoder with intra-span and inter-span information. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1692–1705, Online. Association… [find](https://scholar.google.com/scholar?q=Yijun+Wang%2C+Changzhi+Sun%2C+Yuanbin+Wu%2C+Junchi+Yan%2C+Peng+Gao%2C+and+Guotong+Xie%2E+2020%2E+Pre%2Dtraining+entity+relation+encoder+with+intra%2Dspan+and+inter%2Dspan+information%2E+In+Proceedings+of+the+2020+Conferenc)
11. Sewon Min, Xinxi Lyu, Ari Holtzman, Mikel Artetxe, Mike Lewis, Hannaneh Hajishirzi, and Luke Zettlemoyer. 2022. [Rethinking the role of demonstrations: What makes in-context learning work?](https://md.research.elamur.ai/free/en/papers/min-2022-rethinking-role-demonstrations-what-makes/) In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 11048–11064, Ab…
12. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. 2022. [Chain-of-thought prompting elicits reasoning in large language models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/). Advances in Neural Information Processing Systems, 35:24824–24837.
13. Kaiwen Wei, Yiran Yang, Li Jin, Xian Sun, Zequn Zhang, Jingyuan Zhang, Xiao Li, Linhao Zhang, Jintao Liu, and Guo Zhi. 2023. Guide the many-to-one assignment: Open information extraction via IoU-aware optimal transport. In Proceedings of the 61st Annual Meeting of the Association for Computational L… [find](https://scholar.google.com/scholar?q=Kaiwen+Wei%2C+Yiran+Yang%2C+Li+Jin%2C+Xian+Sun%2C+Zequn+Zhang%2C+Jingyuan+Zhang%2C+Xiao+Li%2C+Linhao+Zhang%2C+Jintao+Liu%2C+and+Guo+Zhi%2E+2023%2E+Guide+the+many%2Dto%2Done+assignment%3A+Open+information+extraction+via+IoU%2Daware)
14. Vlad Niculae, Joonsuk Park, and Claire Cardie. 2017. [Argument mining with structured SVMs and RNNs](https://md.research.elamur.ai/free/en/papers/niculae-2017-argument-mining-structured-svms-rnns/). In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 985–995, Vancouver, Canada. Association for Computational Linguistics.
15. Raquel Mochales Palau and Marie-Francine Moens. 2009. Argumentation mining: the detection, classification and structure of arguments in text. In Proceedings of the 12th international conference on artificial intelligence and law, pages 98–107. [find](https://scholar.google.com/scholar?q=Raquel+Mochales+Palau+and+Marie%2DFrancine+Moens%2E+2009%2E+Argumentation+mining%3A+the+detection%2C+classification+and+structure+of+arguments+in+text%2E+In+Proceedings+of+the+12th+international+conference+on+art)
16. Sun Wei, Mingxiao Li, Jingyuan Sun, Jesse Davis, and Marie-Francine Moens. 2024. DMON: A simple yet effective approach for argument structure learning. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages… [find](https://scholar.google.com/scholar?q=Sun+Wei%2C+Mingxiao+Li%2C+Jingyuan+Sun%2C+Jesse+Davis%2C+and+Marie%2DFrancine+Moens%2E+2024%2E+DMON%3A+A+simple+yet+effective+approach+for+argument+structure+learning%2E+In+Proceedings+of+the+2024+Joint+International+C)
17. Prakash Poudyal, Jaromír Šavelka, Aagje Ieven, Marie Francine Moens, Teresa Goncalves, and Paulo Quaresma. 2020. Echr: legal corpus for argument mining. In Proceedings of the 7th Workshop on Argument Mining, pages 67–75. [find](https://scholar.google.com/scholar?q=Prakash+Poudyal%2C+Jarom%C3%ADr+%C5%A0avelka%2C+Aagje+Ieven%2C+Marie+Francine+Moens%2C+Teresa+Goncalves%2C+and+Paulo+Quaresma%2E+2020%2E+Echr%3A+legal+corpus+for+argument+mining%2E+In+Proceedings+of+the+7th+Workshop+on+Argument+)
18. Adam Wyner, Raquel Mochales-Palau, Marie-Francine Moens, and David Milward. 2010. Approaches to Text Mining Arguments from Legal Cases, pages 60– 79. Springer Berlin Heidelberg, Berlin, Heidelberg. [find](https://scholar.google.com/scholar?q=Adam+Wyner%2C+Raquel+Mochales%2DPalau%2C+Marie%2DFrancine+Moens%2C+and+David+Milward%2E+2010%2E+Approaches+to+Text+Mining+Arguments+from+Legal+Cases%2C+pages+60%E2%80%93+79%2E+Springer+Berlin+Heidelberg%2C+Berlin%2C+Heidelberg%2E)
19. Christian Stab and Iryna Gurevych. 2017. [Parsing argumentation structures in persuasive essays](https://md.research.elamur.ai/free/en/papers/stab-2017-parsing-argumentation-structures-persuasive-essays/). Computational Linguistics, 43(3):619–659.
20. Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Thomas L. Griffiths, Yuan Cao, and Karthik R Narasimhan. 2023. Tree of thoughts: Deliberate problem solving with large language models. In Thirty-seventh Conference on Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Shunyu+Yao%2C+Dian+Yu%2C+Jeffrey+Zhao%2C+Izhak+Shafran%2C+Thomas+L%2E+Griffiths%2C+Yuan+Cao%2C+and+Karthik+R+Narasimhan%2E+2023%2E+Tree+of+thoughts%3A+Deliberate+problem+solving+with+large+language+models%2E+In+Thirty%2Dseve)
21. Nikolaos Stylianou and Ioannis Vlahavas. 2021. Transformed: End-to-end transformers for evidence-based medicine and argument mining in medical literature. Journal of Biomedical Informatics, 117:103767. [find](https://scholar.google.com/scholar?q=Nikolaos+Stylianou+and+Ioannis+Vlahavas%2E+2021%2E+Transformed%3A+End%2Dto%2Dend+transformers+for+evidence%2Dbased+medicine+and+argument+mining+in+medical+literature%2E+Journal+of+Biomedical+Informatics%2C+117%3A103767)
22. Fadi Thabtah, Suhel Hammoud, Firuz Kamalov, and Amanda Gonsalves. 2020. Data imbalance in classification: Experimental evaluation. Information Sciences, 513:429–441. [find](https://scholar.google.com/scholar?q=Fadi+Thabtah%2C+Suhel+Hammoud%2C+Firuz+Kamalov%2C+and+Amanda+Gonsalves%2E+2020%2E+Data+imbalance+in+classification%3A+Experimental+evaluation%2E+Information+Sciences%2C+513%3A429%E2%80%93441%2E)
23. Anar Yeginbergenova and Rodrigo Agerri. 2023. Cross-lingual argument mining in the medical domain. arXiv preprint arXiv:2301.10527. — [arXiv:2301.10527](https://arxiv.org/abs/2301.10527) [find](https://scholar.google.com/scholar?q=Anar+Yeginbergenova+and+Rodrigo+Agerri%2E+2023%2E+Cross%2Dlingual+argument+mining+in+the+medical+domain%2E+arXiv+preprint+arXiv%3A2301%2E10527%2E+%E2%80%94+%5BarXiv%3A2301%2E10527%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2301%2E10527%29)
24. Yu Zeng, Yunzhi Zhuge, Huchuan Lu, Lihe Zhang, Mingyang Qian, and Yizhou Yu. 2019. Multi-source weak supervision for saliency detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6074–6083. [find](https://scholar.google.com/scholar?q=Yu+Zeng%2C+Yunzhi+Zhuge%2C+Huchuan+Lu%2C+Lihe+Zhang%2C+Mingyang+Qian%2C+and+Yizhou+Yu%2E+2019%2E+Multi%2Dsource+weak+supervision+for+saliency+detection%2E+In+Proceedings+of+the+IEEE%2FCVF+conference+on+computer+vision+an)
25. Dawei Zhu, Xiaoyu Shen, Marius Mosbach, Andreas Stephan, and Dietrich Klakow. 2023. Weaker than you think: A critical look at weakly supervised learning. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 14229–14253, Toronto, Ca… [find](https://scholar.google.com/scholar?q=Dawei+Zhu%2C+Xiaoyu+Shen%2C+Marius+Mosbach%2C+Andreas+Stephan%2C+and+Dietrich+Klakow%2E+2023%2E+Weaker+than+you+think%3A+A+critical+look+at+weakly+supervised+learning%2E+In+Proceedings+of+the+61st+Annual+Meeting+of+t)

*Updated 1 November 2026*