Rao, Guozheng, et al. "MGL-CNN: a hierarchical posts representations model for identifying depressed individuals in online forums." IEEE Access 8 (2020): 32395-32403.

This paper introduces two deep learning models, Multi-Gated LeakyReLU CNN (MGL-CNN) and Single-Gated LeakyReLU CNN (SGL-CNN), for detecting depression in online forum posts. MGL-CNN uses hierarchical post-level and user-level operations to capture sentiment from large volumes of user posts, while SGL-CNN varies the number of gated units in the model. Both models were tested on the Reddit Self-reported Depression Diagnosis dataset and the Early Detection of Depression dataset, outperforming existing models in detecting depressive symptoms.