---
title: "Team VYN at SemEval-2026 Task 3: Dimensional Aspect-Based"
description: "The paper presents Team VYN's entry for SemEval-2026 Task 3 (DimABSA 2026), Track A… 3.55% F1. 19 references, cited by 1 paper here."
authors: ["Vishal Thenuwara", "Widanalage De Mel", "Nisansa De Silva"]
year: 2026
venue: "Proceedings of the 20th International Workshop on Semantic Evaluation (2026)"
source: "https://aclanthology.org/2026.semeval-1.223/"
doi: "10.18653/v1/2026.semeval-1.223"
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/thenuwara-2026-team-vyn-semeval-2026-task/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/thenuwara-2026-team-vyn-semeval-2026-task/)

[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/18/) / Team VYN at SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis

# Team VYN at SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis

**Authors:** Vishal Thenuwara, Widanalage De Mel and Nisansa De Silva · **Year:** 2026 · **Venue:** [Proceedings of the 20th International Workshop on Semantic Evaluation (2026)](https://md.research.elamur.ai/free/en/venues/semeval/) ([SemEval 2026](https://md.research.elamur.ai/free/en/venues/semeval/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.semeval-1.223/) · [PDF](https://aclanthology.org/2026.semeval-1.223.pdf) · [DOI 10.18653/v1/2026.semeval-1.223](https://doi.org/10.18653/v1/2026.semeval-1.223) · [OpenAlex](https://openalex.org/W7166824570)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · 1 in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper presents Team VYN's entry for SemEval-2026 Task 3 (DimABSA 2026), Track A English. The task replaces categorical aspect sentiment labels with continuous Valence-Arousal (VA) scores in the range [1, 9]. It covers restaurant and laptop reviews across three subtasks: VA regression for given aspects (DimASR), triplet extraction (DimASTE) and quadruplet prediction with categories (DimASQP).

The authors built three systems:
- **Subtask 1:** a DeBERTa-v3-base regressor with a small MLP head.
- **Subtasks 2 and 3:** a bidirectional MRC-based "Pipeline-DeBERTa" system with several task-specific heads.
- **Subtask 2 alternative:** an adaptation of their span-based DESS model, which combines a BiLSTM, syntactic and semantic GCN channels and a Tree Interaction Network. Its sentiment classifier was swapped for a VA regression head, and training adds a regulariser on batch-level prediction variance.

Results were modest. DESS peaked at 8.22% micro cF1 on the development set, and the system scored below both official LLM baselines. The main lesson is that turning a categorical span-extraction model into a regressor takes more than replacing its output head. The authors argue that architectures designed for joint extraction and continuous regression are needed.

## Key points
- **Data:** 1,448 restaurant and 2,279 laptop training samples (3,727 in total), plus 200 test samples per domain.
- **DESS learning curve:** no valid triplets were produced in epochs 1–2. The model reached 3.55% F1 at epoch 3 and peaked at epoch 5 with 8.22% micro cF1 (precision 8.40%, recall 8.06%).
- **R² for Valence / Arousal:**

  | Subtask | Valence | Arousal |
  |---|---|---|
  | Subtask 1 | 0.589 | −0.349 (worse than a mean baseline) |
  | Subtask 2 | 0.659 | 0.412 |
  | Subtask 3 | 0.681 | 0.423 |
- **Official comparison (restaurant English):** the team scored 1.7978, against 2.1461 for Kimi-K2 Thinking and 2.6427 for Qwen-3 14B, and the authors report this as falling short of both LLM baselines.
- **Constraints and outlook:** training ran on Kaggle T4 GPUs with batch size 4. Suggested next steps include decoupled training, Huber or Smooth L1 loss and binned VA prediction. Code is released at github.com/VishalRepos/DimABSA2026.

*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

- [University of Moratuwa](https://md.research.elamur.ai/free/en/universities/university-of-moratuwa/) — University, Moratuwa, Sri Lanka · 3 works · 456 citations

📚 Shelves (6): [🖥️ 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 · [💭 Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/) · 1,552 works · [💭 Experimental and Cognitive Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/experimental-and-cognitive-psychology/) · 870 works

🧭 Topics (1): [😊 Sentiment & emotion](https://md.research.elamur.ai/free/en/topics/sentiment-emotion/) · 972 works

## Cite

Cite as: Vishal Thenuwara, Widanalage De Mel and Nisansa De Silva. 2026. Team VYN at SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis. Proceedings of the 20th International Workshop on Semantic Evaluation (2026).

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

```bibtex
@inproceedings{thenuwara2026team,
  title = {{Team VYN at SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis}},
  author = {Vishal Thenuwara, Widanalage {De Mel} and Nisansa {De Silva}},
  booktitle = {Proceedings of the 20th International Workshop on Semantic Evaluation (2026)},
  year = {2026},
  doi = {10.18653/v1/2026.semeval-1.223},
  url = {https://doi.org/10.18653/v1/2026.semeval-1.223}
}
```

- [← previous Team UBSE at SemEval-2026 Task 4: Adapting Generalist Embeddings for Narrative Representations](https://md.research.elamur.ai/free/en/papers/marogel-2026-team-ubse-semeval-2026-task/)
- [next → TeamLasse at SemEval-2026 Task 3: A Hybrid Generative-Discriminative Framework for Dimensional Aspect-Based Sentiment Analysis](https://md.research.elamur.ai/free/en/papers/strothe-2026-teamlasse-semeval-2026-task-3/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/18/)

## References

19 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Yomal De Mel and Nisansa de Silva. 2025. GeeSanBhava: Sentiment Tagged Sinhala Music Video Comment Data Set. In International Conference on Computational Collective Intelligence, pages 157–171. Springer. [find](https://scholar.google.com/scholar?q=Yomal+De+Mel+and+Nisansa+de+Silva%2E+2025%2E+GeeSanBhava%3A+Sentiment+Tagged+Sinhala+Music+Video+Comment+Data+Set%2E+In+International+Conference+on+Computational+Collective+Intelligence%2C+pages+157%E2%80%93171%2E+Spring)
2. Sadeep Gunathilaka and Nisansa De Silva. 2022. Aspect-based sentiment analysis on mobile application reviews. In 2022 22nd International Conference on Advances in ICT for Emerging Regions (ICTer), pages 183–188. IEEE. [find](https://scholar.google.com/scholar?q=Sadeep+Gunathilaka+and+Nisansa+De+Silva%2E+2022%2E+Aspect%2Dbased+sentiment+analysis+on+mobile+application+reviews%2E+In+2022+22nd+International+Conference+on+Advances+in+ICT+for+Emerging+Regions+%28ICTer%29%2C+pag)
3. Pengcheng He, Jianfeng Gao, and Weizhu Chen. 2023. [DeBERTaV3: Improving DeBERTa using ELECTRA-style pre-training with gradient-disentangled embedding sharing](https://md.research.elamur.ai/free/en/papers/he-2021-debertav3-improving-deberta-using-electra/). In International Conference on Learning Representations.
4. Dineth Jayakody, Koshila Isuranda, A V A Malkith, Nisansa de Silva, Sachintha Rajith Ponnamperuma, G G N Sandamali, and K L K Sudheera. 2024a. [find](https://scholar.google.com/scholar?q=Dineth+Jayakody%2C+Koshila+Isuranda%2C+A+V+A+Malkith%2C+Nisansa+de+Silva%2C+Sachintha+Rajith+Ponnamperuma%2C+G+G+N+Sandamali%2C+and+K+L+K+Sudheera%2E+2024a%2E)
5. Aspect-Based Sentiment Analysis Techniques: A Comparative Study. In 2024 Moratuwa Engineering Research Conference (MERCon), pages 205–210. IEEE. [find](https://scholar.google.com/scholar?q=Aspect%2DBased+Sentiment+Analysis+Techniques%3A+A+Comparative+Study%2E+In+2024+Moratuwa+Engineering+Research+Conference+%28MERCon%29%2C+pages+205%E2%80%93210%2E+IEEE%2E)
6. Dineth Jayakody, Koshila Isuranda, Ava Malkith, Nisansa de Silva, Sachintha Rajith Ponnamperuma, Ggn Sandamali, Klk Sudheera, and Kash-nika Gimhani Sarathchandra. 2024b. Enhanced aspect-based sentiment analysis with integrated category extraction for instruct-DeBERTa. In Proceedings of the 38th Paci… [find](https://scholar.google.com/scholar?q=Dineth+Jayakody%2C+Koshila+Isuranda%2C+Ava+Malkith%2C+Nisansa+de+Silva%2C+Sachintha+Rajith+Ponnamperuma%2C+Ggn+Sandamali%2C+Klk+Sudheera%2C+and+Kash%2Dnika+Gimhani+Sarathchandra%2E+2024b%2E+Enhanced+aspect%2Dbased+sentimen)
7. Dineth Jayakody, A V A Malkith, Koshila Isuranda, Vishal Thenuwara, Nisansa de Silva, Sachintha Rajith Ponnamperuma, G G N Sandamali, and K L K Sudheera. 2025. Instruct-DeBERTa: A Hybrid Approach for Aspect-based Sentiment Analysis on Textual Reviews. ICTer, 18(2):39–50. [find](https://scholar.google.com/scholar?q=Dineth+Jayakody%2C+A+V+A+Malkith%2C+Koshila+Isuranda%2C+Vishal+Thenuwara%2C+Nisansa+de+Silva%2C+Sachintha+Rajith+Ponnamperuma%2C+G+G+N+Sandamali%2C+and+K+L+K+Sudheera%2E+2025%2E+Instruct%2DDeBERTa%3A+A+Hybrid+Approach+for+)
8. Lung-Hao Lee, Liang-Chih Yu, Natalia Loukashevich, Ilseyar Alimova, Alexander Panchenko, Tzu-Mi Lin, Zhe-Yu Xu, Jian-Yu Zhou, Guangmin Zheng, Jin Wang, et al. 2026. Dimabsa: Building multilingual and multidomain datasets for dimensional aspect-based sentiment analysis. arXiv preprint arXiv:2601.2302… [find](https://scholar.google.com/scholar?q=Lung%2DHao+Lee%2C+Liang%2DChih+Yu%2C+Natalia+Loukashevich%2C+Ilseyar+Alimova%2C+Alexander+Panchenko%2C+Tzu%2DMi+Lin%2C+Zhe%2DYu+Xu%2C+Jian%2DYu+Zhou%2C+Guangmin+Zheng%2C+Jin+Wang%2C+et+al%2E+2026%2E+Dimabsa%3A+Building+multilingual+and+)
9. Chanika Ruchini Mudalige, Dilini Karunarathna, Isanka Rajapaksha, Nisansa de Silva, Gathika Ratnayaka, Amal Shehan Perera, and Ramesh Pathirana. 2020. SigmaLaw-ABSA: Dataset for Aspect-Based Sentiment Analysis in Legal Opinion Texts. In 2020 IEEE 15th International Conference on Industrial and Infor… [find](https://scholar.google.com/scholar?q=Chanika+Ruchini+Mudalige%2C+Dilini+Karunarathna%2C+Isanka+Rajapaksha%2C+Nisansa+de+Silva%2C+Gathika+Ratnayaka%2C+Amal+Shehan+Perera%2C+and+Ramesh+Pathirana%2E+2020%2E+SigmaLaw%2DABSA%3A+Dataset+for+Aspect%2DBased+Sentiment)
10. Maria Pontiki, Dimitris Galanis, John Pavlopoulos, Harris Papageorgiou, Ion Androutsopoulos, and Suresh Manandhar. 2014. [SemEval-2014 task 4: Aspect based sentiment analysis](https://md.research.elamur.ai/free/en/papers/pontiki-2014-semeval-2014-task-4-aspect/). In Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), pages 27–35. Association for Computat…
11. Isanka Rajapaksha, Chanika Ruchini Mudalige, Dilini Karunarathna, Nisansa de Silva, Gathika Rathnayaka, and Amal Shehan Perera. 2020. Rule-Based Approach for Party-Based Sentiment Analysis in Legal Opinion Texts. In 2020 20th International Conference on Advances in ICT for Emerging Regions (ICTer),… [find](https://scholar.google.com/scholar?q=Isanka+Rajapaksha%2C+Chanika+Ruchini+Mudalige%2C+Dilini+Karunarathna%2C+Nisansa+de+Silva%2C+Gathika+Rathnayaka%2C+and+Amal+Shehan+Perera%2E+2020%2E+Rule%2DBased+Approach+for+Party%2DBased+Sentiment+Analysis+in+Legal+Op)
12. Isanka Rajapaksha, Chanika Ruchini Mudalige, Dilini Karunarathna, Nisansa de Silva, Amal Shehan Perera, and Gathika Ratnayaka. 2021a. Sigmalaw PBSA-A Deep Learning Model for Aspect-Based Sentiment Analysis for the Legal Domain. In International Conference on Database and Expert Systems Applications,… [find](https://scholar.google.com/scholar?q=Isanka+Rajapaksha%2C+Chanika+Ruchini+Mudalige%2C+Dilini+Karunarathna%2C+Nisansa+de+Silva%2C+Amal+Shehan+Perera%2C+and+Gathika+Ratnayaka%2E+2021a%2E+Sigmalaw+PBSA%2DA+Deep+Learning+Model+for+Aspect%2DBased+Sentiment+Ana)
13. Isanka Rajapaksha, Chanika Ruchini Mudalige, Dilini Karunarathna, Nisansa de Silva, Gathika Ratnayaka, and Amal Shehan Perera. 2021b. Sigmalaw PBSA-A Deep Learning Approach for Aspect-Based Sentiment Analysis in Legal Opinion Texts. Journal of Data Intelligence, 3(1):101–115. [find](https://scholar.google.com/scholar?q=Isanka+Rajapaksha%2C+Chanika+Ruchini+Mudalige%2C+Dilini+Karunarathna%2C+Nisansa+de+Silva%2C+Gathika+Ratnayaka%2C+and+Amal+Shehan+Perera%2E+2021b%2E+Sigmalaw+PBSA%2DA+Deep+Learning+Approach+for+Aspect%2DBased+Sentiment+)
14. Kevin Scaria, Himanshu Gupta, Siddharth Goyal, Saurabh Sawant, Swaroop Mishra, and Chitta Baral. 2024. InstructABSA: Instruction learning for aspect based sentiment analysis. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human L… [find](https://scholar.google.com/scholar?q=Kevin+Scaria%2C+Himanshu+Gupta%2C+Siddharth+Goyal%2C+Saurabh+Sawant%2C+Swaroop+Mishra%2C+and+Chitta+Baral%2E+2024%2E+InstructABSA%3A+Instruction+learning+for+aspect+based+sentiment+analysis%2E+In+Proceedings+of+the+202)
15. Vishal Thenuwara and Nisansa de Silva. 2025. Dess: Deberta enhanced syntactic-semantic aspect sentiment triplet extraction. In International Conference on Computational Collective Intelligence, pages 185–199. Springer. [find](https://scholar.google.com/scholar?q=Vishal+Thenuwara+and+Nisansa+de+Silva%2E+2025%2E+Dess%3A+Deberta+enhanced+syntactic%2Dsemantic+aspect+sentiment+triplet+extraction%2E+In+International+Conference+on+Computational+Collective+Intelligence%2C+pages+)
16. Zhen Wu, Chengcan Ying, Fei Zhao, Zhifang Fan, Xinyu Dai, and Rui Xia. 2020. Grid tagging scheme for aspect-oriented fine-grained opinion extraction. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 2576–2585. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Zhen+Wu%2C+Chengcan+Ying%2C+Fei+Zhao%2C+Zhifang+Fan%2C+Xinyu+Dai%2C+and+Rui+Xia%2E+2020%2E+Grid+tagging+scheme+for+aspect%2Doriented+fine%2Dgrained+opinion+extraction%2E+In+Findings+of+the+Association+for+Computational+L)
17. Lu Xu, Yew Ken Chia, and Lidong Bing. 2021. [Learning span-level interactions for aspect sentiment triplet extraction](https://md.research.elamur.ai/free/en/papers/xu-2021-learning-span-level-interactions-aspect/). In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long…
18. Liang-Chih Yu, Lung-Hao Lee, et al. 2026. SemEval-2026 task 3: DimABSA – dimensional aspect-based sentiment analysis. In Proceedings of the 20th International Workshop on Semantic Evaluation (SemEval-2026), Vienna, Austria. Association for Computational Linguistics. To appear. [find](https://scholar.google.com/scholar?q=Liang%2DChih+Yu%2C+Lung%2DHao+Lee%2C+et+al%2E+2026%2E+SemEval%2D2026+task+3%3A+DimABSA+%E2%80%93+dimensional+aspect%2Dbased+sentiment+analysis%2E+In+Proceedings+of+the+20th+International+Workshop+on+Semantic+Evaluation+%28SemEval%2D)
19. Wenxuan Zhang, Yang Deng, Xin Li, Yifei Yuan, Lidong Bing, and Wai Lam. 2021. [Aspect sentiment quad prediction as paraphrase generation](https://md.research.elamur.ai/free/en/papers/zhang-2021-aspect-sentiment-quad-prediction-as/). In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 9209– 9219. Association for Computational Linguistics.

*Updated 31 October 2026*