Improving Pre-trained Language Model Sensitivity via Mask Specific losses: A case study on Biomedical NER

Abaho, M, Bollegala, D ORCID: 0000-0003-4476-7003, Leeming, G ORCID: 0000-0002-5554-5302, Joyce, D ORCID: 0000-0002-9433-5340 and Buchan, IE ORCID: 0000-0003-3392-1650 (2024) Improving Pre-trained Language Model Sensitivity via Mask Specific losses: A case study on Biomedical NER In: Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), 2024-6 - 2024-6, Mexico City, Mexico.

Text Learning_Mask_Specific_Losses__DB_.pdf - Author Accepted Manuscript Download (1MB) | Preview

Abstract

Adapting language models (LMs) to novel domains is often achieved through fine-tuning a pre-trained LM (PLM) on domain-specific data. Fine-tuning introduces new knowledge into an LM, enabling it to comprehend and efficiently perform a target domain task. Fine-tuning can however be inadvertently insensitive if it ignores the wide array of disparities (e.g in word meaning) between source and target domains. For instance, words such as chronic and pressure may be treated lightly in social conversations, however, clinically, these words are usually an expression of concern. To address insensitive fine-tuning, we propose Mask Specific Language Modeling (MSLM), an approach that efficiently acquires target domain knowledge by appropriately weighting the importance of domain-specific terms (DS-terms) during fine-tuning. MSLM jointly masks DS-terms and generic words, then learns mask-specific losses by ensuring LMs incur larger penalties for inaccurately predicting DS-terms compared to generic words. Results of our analysis show that MSLM improves LMs sensitivity and detection of DS-terms. We empirically show that an optimal masking rate not only depends on the LM, but also on the dataset and the length of sequences. Our proposed masking strategy outperforms advanced masking strategies such as span- and PMI-based masking.

Item Type:	Conference Item (Unspecified)
Uncontrolled Keywords:	47 Language, Communication and Culture, 4704 Linguistics
Divisions:	Faculty of Health & Life Sciences Faculty of Science & Engineering > School of Electrical Engineering, Electronics and Computer Science Faculty of Health & Life Sciences > Inst. Population Health
Depositing User:	Symplectic Admin
Date Deposited:	07 Jun 2024 07:36
Last Modified:	28 Feb 2026 20:38
DOI:	10.18653/v1/2024.naacl-long.280
Related Websites:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3182080
Disclaimer:	The University of Liverpool is not responsible for content contained on other websites from links within repository metadata. Please contact us if you notice anything that appears incorrect or inappropriate.