User Text Analysis for Mental Health Insights using Transformer-based Models
Bibliographic record
Abstract
The increasing global occurrence of mental health disorders such as anxiety, depression, bipolar disorder, and suicidal ideation highlights the critical need for scalable and automated detection mechanisms. Traditional methods often lack real-time applicability and may exhibit biases due to data imbalance, limiting their effectiveness in early identification. This work addresses these challenges by proposing a Natural Language-based mental health classification system leveraging a fine-tuned Bidirectional Encoder Representations from Transformers (BERT) model to analyze user-generated textual data. A Kaggle mental health dataset comprising 53,043 statements was pre-processed through normalization, and class imbalance was mitigated using RandomOverSampler (ROS). Without ROS, the model showed a strong bias toward the majority classes, leading to suboptimal performance, particularly for underrepresented conditions. Post-ROS implementation, the system achieved notable improvements, with accuracy rising from 87% to 91%, and F1-scores for minority classes such as Bipolar and Suicidal reaching up to 0.99. The final model was integrated into a Streamlit-based web application to enable real-time, accessible mental health screening. Overall, this approach offers a scalable solution for digital mental health monitoring and early intervention by helping identify and support the treatment of individuals showing signs of psychological distress.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.004 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".