Mental distress among Generation Z has emerged as a significant public health concern due to increased exposure to digital environments, academic stress, and social pressures. This survey paper analyzes existing methodologies for early detection of mental distress using machine learning and artificial intelligence techniques. The study reviews various approaches including Natural Language Processing (NLP), sentiment analysis, and deep learning models applied to social media and behavioral datasets. It also explores commonly used datasets, particularly from Kaggle, and evaluates their effectiveness in predictive modeling. The survey identifies limitations in current systems such as data bias, privacy concerns, and lack of real-time adaptability. Furthermore, research gaps and future directions are discussed, emphasizing the need for multimodal data integration and ethical AI deployment.
Thakur et al. (Thu,) studied this question.