Raj, Anuraag and Sharma, Anuraganand and Ali, Zain and Chaudhry, Shonal and Bali, Kavitesh (2024) Depression Detection Using BERT on Social Media Platforms. [Conference Proceedings]
Full text not available from this repository. (Request a copy)Abstract
Depression detection from social media hasattracted significant attention for its potential to offer earlyintervention and support to individuals facing mental healthissues. In this study, we present a comprehensive evaluation ofdeep learning techniques for depression detection, with aspecific focus on leveraging BERT, a powerful NaturalLanguage Processing (NLP) Transformer model. Ourexploration encompasses tailored preprocessing techniques forsocial media text, diverse feature extraction methods, andoptimized model architectures tailored for depression detectiontasks using BERT. Through rigorous experimentation andevaluation, we compare the performance of different BERT-based strategies, considering metrics such as accuracy,efficiency, and scalability. Additionally, we conduct acomparative analysis of labeled and unlabeled data from thesame dataset. For labeled data, we employ BERT directly, whilefor unlabeled data, an autoencoder is utilized following labelremoval. The findings indicate that BERT outperformed othermethods, achieving a high F1-score of 93% on the Redditdataset. BERT achieved an impressive test accuracy of 91.92%,surpassing the Autoencoder model, which attained 84.84%.
| Item Type: | Conference Proceedings |
|---|---|
| Additional Information: | IEEE Xplore page: https://ieeexplore.ieee.org/xpl/conhome/10729761/proceeding |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | School of Information Technology, Engineering, Mathematics and Physics (STEMP) |
| Depositing User: | Anuraganand Sharma |
| Date Deposited: | 30 Jan 2025 02:21 |
| Last Modified: | 30 Jan 2025 02:21 |
| URI: | https://repository.usp.ac.fj/id/eprint/14794 |
Actions (login required)
 |
View Item |