📄 Preprints

These are papers that are currently in preprint form and have not yet been peer-reviewed or published in a conference or journal. Some of them are for a class requirement, while others are for sharing ongoing research work.

Scalable Early Detection of Mental Health Signals in Social Media Text by E. Maminta (October 2025)

Abstract. Early detection of mental health signals in social media text can enable timely interventions and reduce the risk of crises. This paper presents a systematic study of encoder adaptation strategies for automated screening of mental-health indicators in social media posts, with a primary focus on BERT-based encoders. Using a curated dataset labeled across seven clinically informed categories, we compare frozen-encoder classifiers, full encoder fine-tuning, and parameterefficient approaches including low-rank adaptation (LoRA). Through extensive hyperparameter sweeps and controlled ablations we evaluate layer-wise unfreezing schedules, encoder/classifier head learning-rate splits, and regularization schemes. We show that full encoder fine-tuning achieves the highest classification performance with a peak F1-score of 0.83, substantially outperforming frozen and low-rank alternatives while maintaining strong generalization and balanced class sensitivity. Partial fine-tuning and LoRA offer notable compute and memory savings but require careful parameterization to avoid instability at large LoRA ranks. Overall, the findings demonstrate that a fine-tuned BERT encoder can serve as a reliable foundation for early diagnostic support, provided its deployment remains grounded in ethical design and clinical responsibility.