Chinchunmei at WASSA 2024 Empathy and Personality Shared Task: Boosting LLM’s Prediction with Role-play Augmentation and Contrastive Reasoning Calibration

Tian Li, Nicolay Rusnachenko, Huizhi Liang


Abstract
This paper presents the Chinchunmei team’s contributions to the WASSA2024 Shared-Task 1: Empathy Detection and Emotion Classification. We participated in Tracks 1, 2, and 3 to predict empathetic scores based on dialogue, article, and essay content. We choose Llama3-8b-instruct as our base model. We developed three supervised fine-tuning schemes: standard prediction, role-play, and contrastive prediction, along with an innovative scoring calibration method called Contrastive Reasoning Calibration during inference. Pearson Correlation was used as the evaluation metric across all tracks. For Track 1, we achieved 0.43 on the devset and 0.17 on the testset. For Track 2 emotion, empathy, and polarity labels, we obtained 0.64, 0.66, and 0.79 on the devset and 0.61, 0.68, and 0.58 on the testset. For Track 3 empathy and distress labels, we got 0.64 and 0.56 on the devset and 0.33 and 0.35 on the testset.
Anthology ID:
2024.wassa-1.32
Volume:
Proceedings of the 14th Workshop on Computational Approaches to Subjectivity, Sentiment, & Social Media Analysis
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Orphée De Clercq, Valentin Barriere, Jeremy Barnes, Roman Klinger, João Sedoc, Shabnam Tafreshi
Venues:
WASSA | WS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
385–392
Language:
URL:
https://aclanthology.org/2024.wassa-1.32
DOI:
Bibkey:
Cite (ACL):
Tian Li, Nicolay Rusnachenko, and Huizhi Liang. 2024. Chinchunmei at WASSA 2024 Empathy and Personality Shared Task: Boosting LLM’s Prediction with Role-play Augmentation and Contrastive Reasoning Calibration. In Proceedings of the 14th Workshop on Computational Approaches to Subjectivity, Sentiment, & Social Media Analysis, pages 385–392, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Chinchunmei at WASSA 2024 Empathy and Personality Shared Task: Boosting LLM’s Prediction with Role-play Augmentation and Contrastive Reasoning Calibration (Li et al., WASSA-WS 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.wassa-1.32.pdf