CogAI@SMM4H 2024: Leveraging BERT-based Ensemble Models for Classifying Tweets on Developmental Disorders

AbstractThis paper presents our work for the Task 5 of the Social Media Mining for Health Applications 2024 Shared Task - Binary classification of English tweets reporting children’s medical disorders. In this paper, we present and compare multiple approaches for automatically classifying tweets from parents based on whether they mention having a child with attention-deficit/hyperactivity disorder (ADHD), autism spectrum disorders (ASD), delayed speech, or asthma. We use ensemble of various BERT-based models trained on provided dataset that yields an F1 score of 0.901 on the test data.