SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis

SemEval-2026 Task 3 introduces DimABSA and DimStance tasks, advancing NLP research in sentiment analysis and stance detection across 9 languages and 6 domains.

A new shared task, SemEval-2026 Task 3, is announced, focusing on Dimensional Aspect-Based Sentiment Analysis (DimABSA) and Dimensional Stance Analysis (DimStance).

The task aims to bridge the gap between traditional Aspect-Based Sentiment Analysis (ABSA) and dimensional sentiment analysis, which represents sentiment along fine-grained, real-valued dimensions of valence and arousal.

• DimABSA integrates dimensional sentiment analysis into the traditional ABSA framework.
• DimStance reformulates stance detection under the ABSA schema in the valence-arousal space.

The task includes two tracks: Track A (DimABSA) and Track B (DimStance), with subtasks for each.

Data is provided in 9 languages and 6 application domains.

For more details and to participate, visit the task website, register on Codabench, and join the Discord community for discussion.

Tags: Dimensional Aspect-Based Sentiment Analysis, DimABSA, SemEval-2026 Task 3, NLP research, sentiment analysis, stance detection, valence-arousal representation