---
title: "Mitigating Degree Bias in Hypergraphs via Attribute-as-Structure"
description: "The authors argue that the usual remedy fails because it is a feature-level fix. 32 references."
authors: ["Ryusei Nishide", "Makoto Miwa"]
year: 2026
venue: "Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.eacl-long.81/"
doi: "10.18653/v1/2026.eacl-long.81"
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/nishide-2026-mitigating-degree-bias-hypergraphs-attribute/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/nishide-2026-mitigating-degree-bias-hypergraphs-attribute/)

[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/36/) / Mitigating Degree Bias in Hypergraphs via Attribute-as-Structure Approach

# Mitigating Degree Bias in Hypergraphs via Attribute-as-Structure Approach

**Authors:** Ryusei Nishide and [Makoto Miwa](https://md.research.elamur.ai/free/en/authors/makoto-miwa/) · **Year:** 2026 · **Venue:** [Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/eacl/) ([EACL 2026](https://md.research.elamur.ai/free/en/venues/eacl/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.eacl-long.81/) · [PDF](https://aclanthology.org/2026.eacl-long.81.pdf) · [DOI 10.18653/v1/2026.eacl-long.81](https://doi.org/10.18653/v1/2026.eacl-long.81) · [OpenAlex](https://openalex.org/W7140125074)

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

## Summary
Hypergraph models that learn node representations by aggregating over neighbours perform poorly on low-degree nodes. This is known as degree bias: weakly connected nodes have too little structural context to draw from. The authors argue that the usual remedy fails because it is a feature-level fix. That remedy appends textual attributes such as titles, abstracts or keywords to node features, which adds no new routes for information flow. Their "attribute-as-structure" alternative makes those attributes separate node types inside a heterogeneous hypergraph, so sparse entities gain extra aggregation paths.

The framework has two main parts:
- **Encoder.** It uses learnable type, role and edge-type embeddings. It applies two attention mechanisms: self-attention among entities and cross-type attention between entities and attributes. Attribute nodes are initialised from pre-trained encoders such as BERT.
- **Training objective.** A TriCL-based structural loss is combined with Attribute-Attentive Contrastive Learning (AACL). AACL adaptively weights attributes and aligns entities with their attribute nodes.

On node classification over citation hypergraphs (Cora-C, Cora-A, PubMed), the method gives steady accuracy gains, and the gains are largest for sparsely connected nodes. The work suggests that changing topology, rather than enriching features, is the more effective lever against degree bias.

## Key points
- **Headline results:** accuracy improves by up to +1.64% overall and up to +2.10% for structurally sparse nodes. The largest benefit is for nodes with only 1–2 hyperedges. Very dense nodes (≥21 connections) show a slight drop.
- **Comparison with attribute-as-feature baselines on Cora-C / Cora-A:**
  - vs. HyperBERT: +1.69% / +2.15%
  - vs. the authors' own base model with concatenated title–abstract features: +0.88% / +0.86%
- **Cost:** attribute nodes roughly triple the node count, but inference time grows only about 1.5×. On PubMed, for example, it rises from 30 ms to 48 ms.
- **Without attribute nodes:** the base encoder is competitive with or better than existing methods across 10 benchmark hypergraph datasets.
- **Limitations noted by the authors:** scalability overhead on very large graphs, support for text-only attributes, and the need for dataset-specific tuning of the structure/attribute balance.

*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

- [Toyota Technological Institute](https://md.research.elamur.ai/free/en/universities/toyota-technological-institute/) — University, Nagoya, Japan · 9 works · 3,330 citations

📚 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 · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 2,869 works · [⚛️ Physics and Astronomy](https://md.research.elamur.ai/free/en/shelves/physics-and-astronomy/) · 44 works · [⚛️ Statistical and Nonlinear Physics](https://md.research.elamur.ai/free/en/shelves/physics-and-astronomy/statistical-and-nonlinear-physics/) · 40 works

## Cite

Cite as: Ryusei Nishide and Makoto Miwa. 2026. Mitigating Degree Bias in Hypergraphs via Attribute-as-Structure Approach. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers).

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

```bibtex
@inproceedings{nishide2026mitigating,
  title = {{Mitigating Degree Bias in Hypergraphs via Attribute-as-Structure Approach}},
  author = {Ryusei Nishide and Makoto Miwa},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-long.81},
  url = {https://doi.org/10.18653/v1/2026.eacl-long.81}
}
```

- [← previous Mitigating Cultural Bias in LLMs via Multi-Agent Cultural Debate](https://md.research.elamur.ai/free/en/papers/tan-2026-mitigating-cultural-bias-llms-multi/)
- [next → Mitigating Hallucinations in VLMs: Enhancing Visual Attention via Head-Wise Perturbation](https://md.research.elamur.ai/free/en/papers/wang-2026-mitigating-hallucinations-vlms-enhancing-visual/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/36/)

## References

32 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Alessia Antelmi, Gennaro Cordasco, Mirko Polato, Vittorio Scarano, Carmine Spagnuolo, and Dingqi Yang. 2024. A survey on hypergraph representation learning. ACM Comput. Surv., 56(1):24:1–24:38. [find](https://scholar.google.com/scholar?q=Alessia+Antelmi%2C+Gennaro+Cordasco%2C+Mirko+Polato%2C+Vittorio+Scarano%2C+Carmine+Spagnuolo%2C+and+Dingqi+Yang%2E+2024%2E+A+survey+on+hypergraph+representation+learning%2E+ACM+Comput%2E+Surv%2E%2C+56%281%29%3A24%3A1%E2%80%9324%3A38%2E)
2. Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E. Hinton. 2016. [Layer normalization](https://md.research.elamur.ai/free/en/papers/ba-2016-layer-normalization/).
3. Adrián Bazaga, Pietro Lio, and Gos Micklem. 2024. HyperBERT: Mixing hypergraph-aware layers with language models for node classification on text-attributed hypergraphs. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 9181–9193, Miami, Florida, USA. Association for Com… [find](https://scholar.google.com/scholar?q=Adri%C3%A1n+Bazaga%2C+Pietro+Lio%2C+and+Gos+Micklem%2E+2024%2E+HyperBERT%3A+Mixing+hypergraph%2Daware+layers+with+language+models+for+node+classification+on+text%2Dattributed+hypergraphs%2E+In+Findings+of+the+Association+)
4. Ying Chen, Jinbo Bi, and James Z. Wang. 2003. A visual parts-based object recognition system. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). [find](https://scholar.google.com/scholar?q=Ying+Chen%2C+Jinbo+Bi%2C+and+James+Z%2E+Wang%2E+2003%2E+A+visual+parts%2Dbased+object+recognition+system%2E+In+Proceedings+of+the+IEEE+Conference+on+Computer+Vision+and+Pattern+Recognition+%28CVPR%29%2E)
5. Zhixuan Chu, Yan Wang, Qing Cui, Longfei Li, Wenqing Chen, Zhan Qin, and Kui Ren. 2024. Llm-guided multi-view hypergraph learning for human-centric explainable recommendation. Preprint, arXiv:2401.08217. — [arXiv:2401.08217](https://arxiv.org/abs/2401.08217) [find](https://scholar.google.com/scholar?q=Zhixuan+Chu%2C+Yan+Wang%2C+Qing+Cui%2C+Longfei+Li%2C+Wenqing+Chen%2C+Zhan+Qin%2C+and+Kui+Ren%2E+2024%2E+Llm%2Dguided+multi%2Dview+hypergraph+learning+for+human%2Dcentric+explainable+recommendation%2E+Preprint%2C+arXiv%3A2401%2E082)
6. Djork-Arné Clevert, Thomas Unterthiner, and Sepp Hochreiter. 2016. [Fast and accurate deep network learning by exponential linear units (elus)](https://md.research.elamur.ai/free/en/papers/clevert-2015-fast-accurate-deep-network-learning/). In 4th International Conference on Learning Representations, ICLR 2016, San Juan, Puerto Rico, May 2-4, 2016, Conference Track Proceedings.
7. Yihe Dong, Will Sawin, and Yoshua Bengio. 2020. Hnhn: Hypergraph networks with hyperedge neurons. [find](https://scholar.google.com/scholar?q=Yihe+Dong%2C+Will+Sawin%2C+and+Yoshua+Bengio%2E+2020%2E+Hnhn%3A+Hypergraph+networks+with+hyperedge+neurons%2E)
8. Dheeru Dua and Casey Graff. 2017. UCI machine learning repository. [find](https://scholar.google.com/scholar?q=Dheeru+Dua+and+Casey+Graff%2E+2017%2E+UCI+machine+learning+repository%2E)
9. Yifan Feng, Haoxuan You, Zizhao Zhang, Rongrong Ji, and Yue Gao. 2019. Hypergraph neural networks. In Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence and Thirty-First Innovative Applications of Artificial Intelligence Conference and Ninth AAAI Symposium on Educational Adva… [find](https://scholar.google.com/scholar?q=Yifan+Feng%2C+Haoxuan+You%2C+Zizhao+Zhang%2C+Rongrong+Ji%2C+and+Yue+Gao%2E+2019%2E+Hypergraph+neural+networks%2E+In+Proceedings+of+the+Thirty%2DThird+AAAI+Conference+on+Artificial+Intelligence+and+Thirty%2DFirst+Innova)
10. Yifan Feng, Zizhao Zhang, Xibin Zhao, Rongrong Ji, and Yue Gao. 2018. Gvcnn: Group-view convolutional neural networks for 3d shape recognition. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 264–272. [find](https://scholar.google.com/scholar?q=Yifan+Feng%2C+Zizhao+Zhang%2C+Xibin+Zhao%2C+Rongrong+Ji%2C+and+Yue+Gao%2E+2018%2E+Gvcnn%3A+Group%2Dview+convolutional+neural+networks+for+3d+shape+recognition%2E+In+2018+IEEE%2FCVF+Conference+on+Computer+Vision+and+Patte)
11. C. Lee Giles, Kurt D. Bollacker, and Steve Lawrence. 1998. Citeseer: an automatic citation indexing system. In Proceedings of the Third ACM Conference on Digital Libraries, DL ’98, page 89–98. Association for Computing Machinery. [find](https://scholar.google.com/scholar?q=C%2E+Lee+Giles%2C+Kurt+D%2E+Bollacker%2C+and+Steve+Lawrence%2E+1998%2E+Citeseer%3A+an+automatic+citation+indexing+system%2E+In+Proceedings+of+the+Third+ACM+Conference+on+Digital+Libraries%2C+DL+%E2%80%9998%2C+page+89%E2%80%9398%2E+Associa)
12. Yu Gu, Robert Tinn, Hao Cheng, Michael Lucas, Naoto Usuyama, Xiaodong Liu, Tristan Naumann, Jianfeng Gao, and Hoifung Poon. 2021. [Domain-specific language model pretraining for biomedical natural language processing](https://md.research.elamur.ai/free/en/papers/gu-2020-domain-specific-language-model-pretraining/). ACM Trans. Comput. Healthcare, 3(1).
13. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. [Deep residual learning for image recognition](https://md.research.elamur.ai/free/en/papers/he-2015-deep-residual-learning-image-recognition/). In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
14. Xiaoxin He, Xavier Bresson, Thomas Laurent, Adam Perold, Yann LeCun, and Bryan Hooi. 2024. Harnessing explanations: LLM-to-LM interpreter for enhanced text-attributed graph representation learning. In The Twelfth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Xiaoxin+He%2C+Xavier+Bresson%2C+Thomas+Laurent%2C+Adam+Perold%2C+Yann+LeCun%2C+and+Bryan+Hooi%2E+2024%2E+Harnessing+explanations%3A+LLM%2Dto%2DLM+interpreter+for+enhanced+text%2Dattributed+graph+representation+learning%2E+In)
15. Van Thuy Hoang, Hyeon-Ju Jeon, and O-Joun Lee. 2025. Mitigating degree bias in graph representation learning with learnable structural augmentation and structural self-attention. IEEE Transactions on Network Science and Engineering, 12(5):3656–3670. [find](https://scholar.google.com/scholar?q=Van+Thuy+Hoang%2C+Hyeon%2DJu+Jeon%2C+and+O%2DJoun+Lee%2E+2025%2E+Mitigating+degree+bias+in+graph+representation+learning+with+learnable+structural+augmentation+and+structural+self%2Dattention%2E+IEEE+Transactions+on+)
16. Jing Huang and Jie Yang. 2021. Unignn: a unified framework for graph and hypergraph neural networks. In Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, IJCAI-21. [find](https://scholar.google.com/scholar?q=Jing+Huang+and+Jie+Yang%2E+2021%2E+Unignn%3A+a+unified+framework+for+graph+and+hypergraph+neural+networks%2E+In+Proceedings+of+the+Thirtieth+International+Joint+Conference+on+Artificial+Intelligence%2C+IJCAI%2D21)
17. Mingxuan Ju, Tong Zhao, Wenhao Yu, Neil Shah, and Yanfang Ye. 2023. Graphpatcher: Mitigating degree bias for graph neural networks via test-time augmentation. In Thirty-seventh Conference on Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Mingxuan+Ju%2C+Tong+Zhao%2C+Wenhao+Yu%2C+Neil+Shah%2C+and+Yanfang+Ye%2E+2023%2E+Graphpatcher%3A+Mitigating+degree+bias+for+graph+neural+networks+via+test%2Dtime+augmentation%2E+In+Thirty%2Dseventh+Conference+on+Neural+In)
18. Sunwoo Kim, Shinhwan Kang, Fanchen Bu, Soo Yong Lee, Jaemin Yoo, and Kijung Shin. 2024. Hypeboy: Generative self-supervised representation learning on hypergraphs. In The Twelfth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Sunwoo+Kim%2C+Shinhwan+Kang%2C+Fanchen+Bu%2C+Soo+Yong+Lee%2C+Jaemin+Yoo%2C+and+Kijung+Shin%2E+2024%2E+Hypeboy%3A+Generative+self%2Dsupervised+representation+learning+on+hypergraphs%2E+In+The+Twelfth+International+Confere)
19. Ken Lang. 1995. Newsweeder: Learning to filter netnews. pages 331–339. [find](https://scholar.google.com/scholar?q=Ken+Lang%2E+1995%2E+Newsweeder%3A+Learning+to+filter+netnews%2E+pages+331%E2%80%93339%2E)
20. Dongjin Lee and Kijung Shin. 2023. I’m me, we’re us, and i’m us: Tri-directional contrastive learning on hypergraphs. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 8456– 8464. [find](https://scholar.google.com/scholar?q=Dongjin+Lee+and+Kijung+Shin%2E+2023%2E+I%E2%80%99m+me%2C+we%E2%80%99re+us%2C+and+i%E2%80%99m+us%3A+Tri%2Ddirectional+contrastive+learning+on+hypergraphs%2E+In+Proceedings+of+the+AAAI+Conference+on+Artificial+Intelligence%2C+volume+37%2C+pages)
21. Liyuan Liu, Haoming Jiang, Pengcheng He, Weizhu Chen, Xiaodong Liu, Jianfeng Gao, and Jiawei Han. 2020. On the variance of the adaptive learning rate and beyond. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Liyuan+Liu%2C+Haoming+Jiang%2C+Pengcheng+He%2C+Weizhu+Chen%2C+Xiaodong+Liu%2C+Jianfeng+Gao%2C+and+Jiawei+Han%2E+2020%2E+On+the+variance+of+the+adaptive+learning+rate+and+beyond%2E+In+International+Conference+on+Learnin)
22. Zexi Liu, Bohan Tang, Ziyuan Ye, Xiaowen Dong, Siheng Chen, and Yanfeng Wang. 2024. Hypergraph transformer for semi-supervised classification. [find](https://scholar.google.com/scholar?q=Zexi+Liu%2C+Bohan+Tang%2C+Ziyuan+Ye%2C+Xiaowen+Dong%2C+Siheng+Chen%2C+and+Yanfeng+Wang%2E+2024%2E+Hypergraph+transformer+for+semi%2Dsupervised+classification%2E)
23. Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Ilya+Loshchilov+and+Frank+Hutter%2E+2019%2E+Decoupled+weight+decay+regularization%2E+In+International+Conference+on+Learning+Representations%2E)
24. Andrew Kachites McCallum, Kamal Nigam, Jason Rennie, and Kristie Seymore. 2000. Automating the construction of internet portals with machine learning. Information Retrieval, 3(2):127–163. [find](https://scholar.google.com/scholar?q=Andrew+Kachites+McCallum%2C+Kamal+Nigam%2C+Jason+Rennie%2C+and+Kristie+Seymore%2E+2000%2E+Automating+the+construction+of+internet+portals+with+machine+learning%2E+Information+Retrieval%2C+3%282%29%3A127%E2%80%93163%2E)
25. Nils Reimers and Iryna Gurevych. 2019. [Sentence-bert: Sentence embeddings using siamese bert-networks](https://md.research.elamur.ai/free/en/papers/reimers-2019-sentence-bert-sentence-embeddings-siamese/). In Conference on Empirical Methods in Natural Language Processing.
26. Prithviraj Sen, Galileo Namata, Mustafa Bilgic, Lise Getoor, Brian Gallagher, and Tina Eliassi-Rad. 2008. Collective classification in network data. AI Mag., 29(3):93–106. [find](https://scholar.google.com/scholar?q=Prithviraj+Sen%2C+Galileo+Namata%2C+Mustafa+Bilgic%2C+Lise+Getoor%2C+Brian+Gallagher%2C+and+Tina+Eliassi%2DRad%2E+2008%2E+Collective+classification+in+network+data%2E+AI+Mag%2E%2C+29%283%29%3A93%E2%80%93106%2E)
27. Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever, and Ruslan Salakhutdinov. 2014. Dropout: A simple way to prevent neural networks from overfitting. Journal of Machine Learning Research, 15(56):1929–1958. [find](https://scholar.google.com/scholar?q=Nitish+Srivastava%2C+Geoffrey+Hinton%2C+Alex+Krizhevsky%2C+Ilya+Sutskever%2C+and+Ruslan+Salakhutdinov%2E+2014%2E+Dropout%3A+A+simple+way+to+prevent+neural+networks+from+overfitting%2E+Journal+of+Machine+Learning+Rese)
28. Hang Su, Subhransu Maji, Evangelos Kalogerakis, and Erik Learned-Miller. 2015. Multi-view Convolutional Neural Networks for 3D Shape Recognition . In 2015 IEEE International Conference on Computer Vision (ICCV), pages 945–953, Los Alamitos, CA, USA. IEEE Computer Society. [find](https://scholar.google.com/scholar?q=Hang+Su%2C+Subhransu+Maji%2C+Evangelos+Kalogerakis%2C+and+Erik+Learned%2DMiller%2E+2015%2E+Multi%2Dview+Convolutional+Neural+Networks+for+3D+Shape+Recognition+%2E+In+2015+IEEE+International+Conference+on+Computer+Vis)
29. Arjun Subramonian, Jian Kang, and Yizhou Sun. 2024. Theoretical and empirical insights into the origins of degree bias in graph neural networks. In The Thirty-eighth Annual Conference on Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Arjun+Subramonian%2C+Jian+Kang%2C+and+Yizhou+Sun%2E+2024%2E+Theoretical+and+empirical+insights+into+the+origins+of+degree+bias+in+graph+neural+networks%2E+In+The+Thirty%2Deighth+Annual+Conference+on+Neural+Inform)
30. Qiaoyu Tan, Xin Zhang, Xiao Huang, Hao Chen, Jundong Li, and Xia Hu. 2024. Collaborative graph neural networks for attributed network embedding. [find](https://scholar.google.com/scholar?q=Qiaoyu+Tan%2C+Xin+Zhang%2C+Xiao+Huang%2C+Hao+Chen%2C+Jundong+Li%2C+and+Xia+Hu%2E+2024%2E+Collaborative+graph+neural+networks+for+attributed+network+embedding%2E)
31. Yanran Tang, Ruihong Qiu, Yilun Liu, Xue Li, and Zi Huang. 2024. Casegnn: Graph neural networks for legal case retrieval with text-attributed graphs. [find](https://scholar.google.com/scholar?q=Yanran+Tang%2C+Ruihong+Qiu%2C+Yilun+Liu%2C+Xue+Li%2C+and+Zi+Huang%2E+2024%2E+Casegnn%3A+Graph+neural+networks+for+legal+case+retrieval+with+text%2Dattributed+graphs%2E)
32. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. [Attention is all you need](https://md.research.elamur.ai/free/en/papers/vaswani-2017-attention-all-need/). In Proceedings of the 31st International

*Updated 1 November 2026*