SCUNLP at ROCLING-2025 Shared Task: Systematic Guideline Refinement for Continuous Value Prediction with Outlier-Driven LLM Feedback

Hong Rui Pan, Jheng Long Wu


Abstract
Regression-based prediction is widely applied to continuous outputs, such as emotion dimension estimation. However, traditional methods struggle to handle unclear annotation standards and ambiguous cases. To address this challenge, we propose a dual-layer agent-executor framework, where the agent is responsible for constructing and refining guidelines, while the executor applies these guidelines to annotate large-scale data. Notably, we introduce a novel refinement mechanism that can detect outlier instances and provide feedback to the agent for guideline revision, thereby achieving iterative improvement. We applied this method to the ROCLING 2025 shared task for predicting valence-arousal (VA) values in medical self-reflection texts. Compared to the unmodified version, the outlier-driven configuration effectively reduced MAE for both V/A, with A-MAE significantly decreased by 7.7%. The final valence-MAE was 0.51 and arousal-MAE was 0.87, ranking fourth.
Anthology ID:
2025.rocling-main.47
Volume:
Proceedings of the 37th Conference on Computational Linguistics and Speech Processing (ROCLING 2025)
Month:
November
Year:
2025
Address:
National Taiwan University, Taipei City, Taiwan
Editors:
Kai-Wei Chang, Ke-Han Lu, Chih-Kai Yang, Zhi-Rui Tam, Wen-Yu Chang, Chung-Che Wang
Venue:
ROCLING
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
418–426
Language:
URL:
https://aclanthology.org/2025.rocling-main.47/
DOI:
Bibkey:
Cite (ACL):
Hong Rui Pan and Jheng Long Wu. 2025. SCUNLP at ROCLING-2025 Shared Task: Systematic Guideline Refinement for Continuous Value Prediction with Outlier-Driven LLM Feedback. In Proceedings of the 37th Conference on Computational Linguistics and Speech Processing (ROCLING 2025), pages 418–426, National Taiwan University, Taipei City, Taiwan. Association for Computational Linguistics.
Cite (Informal):
SCUNLP at ROCLING-2025 Shared Task: Systematic Guideline Refinement for Continuous Value Prediction with Outlier-Driven LLM Feedback (Pan & Wu, ROCLING 2025)
Copy Citation:
PDF:
https://aclanthology.org/2025.rocling-main.47.pdf