This paper describes our system for the DimABSA 2026 Shared Task (SemEval-2026 Task 3), Track A, covering all three subtasks. We develop two complementary approaches: (1) DESS (Thenuwara and de Silva, 2025), an adaptation of our span-based extraction model incorporating dual-channel GCNs and a valenceâÂÂarousal (VA) regression head.
Keywords: Machine Learning / Deep Learning | Natural Language Processing | Aspect Sentiment Triple Extraction | Aspect-based Sentiment Analysis | Information Extraction | Sentiment Analysis |
@inproceedings{thenuwara2026team,
title={Team VYN at SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis},
author={Thenuwara, Vishal and De Mel, W M Yomal and De Silva, Nisansa},
booktitle={Proceedings of the 20th International Workshop on Semantic Evaluation (2026)},
pages={1755--1764},
year={2026},
abstract={This paper describes our system for the DimABSA 2026 Shared Task (SemEval-2026 Task 3), Track A, covering all three subtasks. We develop two complementary approaches: (1) DESS (Thenuwara and de Silva, 2025), an adaptation of our span-based extraction model incorporating dual-channel GCNs and a valenceâÂÂarousal (VA) regression head.},
doi={10.18653/v1/2026.semeval-1.223},
misc={https://aclanthology.org/2026.semeval-1.223.pdf,ML:NLP,https://goo.gl/iY6aTr}
}
ActivePI: Nisansa de Silva
We propose to advance aspect-based sentiment analysis (ABSA) by developing enhanced models, datasets, and evaluation methods that improve the accuracy and applicability of sentiment classification across multiple domains.