Calibrating Student Models for Emotion-related Tasks

Mahshid Hosseini, Cornelia Caragea


Abstract
Knowledge Distillation (KD) is an effective method to transfer knowledge from one network (a.k.a. teacher) to another (a.k.a. student). In this paper, we study KD on the emotion-related tasks from a new perspective: calibration. We further explore the impact of the mixup data augmentation technique on the distillation objective and propose to use a simple yet effective mixup method informed by training dynamics for calibrating the student models. Underpinned by the regularization impact of the mixup process by providing better training signals to the student models using training dynamics, our proposed mixup strategy gradually enhances the student model’s calibration while effectively improving its performance. We evaluate the calibration of pre-trained language models through knowledge distillation over three tasks of emotion detection, sentiment analysis, and empathy detection. By conducting extensive experiments on different datasets, with both in-domain and out-of-domain test sets, we demonstrate that student models distilled from teacher models trained using our proposed mixup method obtained the lowest Expected Calibration Errors (ECEs) and best performance on both in-domain and out-of-domain test sets.
Anthology ID:
2022.emnlp-main.629
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
9266–9278
Language:
URL:
https://aclanthology.org/2022.emnlp-main.629
DOI:
10.18653/v1/2022.emnlp-main.629
Bibkey:
Cite (ACL):
Mahshid Hosseini and Cornelia Caragea. 2022. Calibrating Student Models for Emotion-related Tasks. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 9266–9278, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Calibrating Student Models for Emotion-related Tasks (Hosseini & Caragea, EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.629.pdf