Shayona@SMM4H23: COVID-19 Self diagnosis classification using BERT and LightGBM models

The research paper presents the work of Team Shayona on two tasks of SMMH4-23. Task 1 involved the binary classification of English tweets self-reporting a COVID-19 diagnosis, and Task 4 was a binary classification of English Reddit posts self-reporting a social anxiety disorder diagnosis. The team used a combination of the Transformer model (BERT) and the LightGBM model for both tasks. Their approach achieved the highest f1-score of 0.94 in Task 1. The paper also discusses the potential improvements for both tasks.

Publication date: 4 Jan 2024
Project Page: https://arxiv.org/abs/2401.02158v1
Paper: https://arxiv.org/pdf/2401.02158