Workshop Proceedings of the 19th International AAAI Conference on Web and Social Media
Workshop: #SMM4H-HeaRD 2025: Joint 10th Social Media Mining for Health and Health Real-World Data Workshop and Shared Tasks
DOI: 10.36190/2025.60Published:
2025-06-05
BioNLP1 at #SMM4H-HeaRD 2025: Stacked Representations for Detection of Vaccine Adverse Event Mentions in Social Media Text
This paper presents the system developed by the BioNLP1 team for Task 6 of the 10th edition of the Social Media Mining for Health (SMM4H) 2025. The task involves the binary classification problem of identifying vaccine adverse event mentions (VAEM) in shingles-related Reddit posts. Our approach leverages a stacking method that combines TF-IDF features with sentence embeddings from a fine-tuned transformer model (RoBERTa-large) to classify the posts. The proposed system achieved a high F1 score of 0.97 on the validation set and the highest F1 score of 0.96 on the test set, with scores computed on the positive class.