UATTA-EB: Uncertainty-Aware Test-Time Augmented Ensemble of BERTs for Classifying Common Mental Illnesses on Social Media Posts
This addresses the need for more accurate mental health screening tools for individuals using online platforms, though it appears incremental as it builds on existing BERT and ensembling methods.
The paper tackled the problem of unreliable and overconfident predictions in deep learning models for classifying mental illnesses from social media posts, proposing UATTA-EB to produce reliable and well-calibrated predictions for six types of mental illnesses using Reddit data.
Given the current state of the world, because of existing situations around the world, millions of people suffering from mental illnesses feel isolated and unable to receive help in person. Psychological studies have shown that our state of mind can manifest itself in the linguistic features we use to communicate. People have increasingly turned to online platforms to express themselves and seek help with their conditions. Deep learning methods have been commonly used to identify and analyze mental health conditions from various sources of information, including social media. Still, they face challenges, including a lack of reliability and overconfidence in predictions resulting in the poor calibration of the models. To solve these issues, We propose UATTA-EB: Uncertainty-Aware Test-Time Augmented Ensembling of BERTs for producing reliable and well-calibrated predictions to classify six possible types of mental illnesses- None, Depression, Anxiety, Bipolar Disorder, ADHD, and PTSD by analyzing unstructured user data on Reddit.