Abstract
This article presents a system description for our work as part of Task 5 of the SMM4H-HeaRD 2026 workshop. We fine-tune pretrained BERT and BiomedBERT models and further enhance them using custom feature augmentation techniques. Incorporating these engineered features results in improved performance, with the best model achieving a validation F1 score of 0.8419 and an evaluation phase F1 score of 0.753.- Anthology ID:
- 2026.smm4h-1.35
- Volume:
- Proceedings of the 11th Social Media Mining for Health Research and Applications (SMM4H-HeaRD 2026) Workshop and Shared Tasks
- Month:
- July
- Year:
- 2026
- Address:
- San Diego, United States
- Editors:
- Guillermo Lopez-Garcia, Graciela Gonzalez-Hernandez
- Venues:
- SMM4H | WS
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 222–224
- Language:
- URL:
- https://preview.aclanthology.org/acl-awards/2026.smm4h-1.35/
- DOI:
- 10.18653/v1/2026.smm4h-1.35
- Cite (ACL):
- Aakarsh Bansal, Abhishek Srinivas, and Sowmya Kamath S.. 2026. HALELab-NITK at #SMM4H-HeaRD2026: Inclusion of Feature Engineering for Detection of Patient Metadata in SARS-CoV2 Sequencing Articles. In Proceedings of the 11th Social Media Mining for Health Research and Applications (SMM4H-HeaRD 2026) Workshop and Shared Tasks, pages 222–224, San Diego, United States. Association for Computational Linguistics.
- Cite (Informal):
- HALELab-NITK at #SMM4H-HeaRD2026: Inclusion of Feature Engineering for Detection of Patient Metadata in SARS-CoV2 Sequencing Articles (Bansal et al., SMM4H 2026)
- PDF:
- https://preview.aclanthology.org/acl-awards/2026.smm4h-1.35.pdf