Deep learning based mental health/status interpretation

• Main Author: Teo, Guang Xiang
• Other Authors: Vidya Sudarshan
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/171965
• Institution: Nanyang Technological University

Detecting mental health disorders through analysis of social media activity is a challenging yet crucial task, particularly in terms of early intervention for individuals experiencing mental health issues. This study introduces an approach to interpreting mental health conditions by employing Deep Learning models on Reddit posts. The research utilized deep learning models to examine and classify posts that are associated with mental health disorders. The primary dataset consisted of Reddit posts, with a specific focus on identifying posts related to depression. Moreover, the dataset was also utilized to categorize posts pertaining to various other mental disorders. The study implemented a two-stage classifier to facilitate effective analysis. The initial stage involved filtering out posts that were not relevant to mental disorders, while the subsequent stage focused on categorizing the remaining posts into specific mental disorders. This innovative two- stage approach offers a fresh perspective on utilizing social media data for mental health analysis and has the potential to make significant contributions to the field of digital mental health.