Reproduce “CapsF: Capsule Fusion for Extracting psychiatric stressors for suicide from twitter” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Mohammad Ali Dadgostarnia, Ramin Mousa, Saba Hesaraki
A deep learning system named CapsF identifies suicide-related psychiatric stressors within Persian tweets by integrating textual features from a Bi-GRUCapsule network with auxiliary emotional and user-centric data. The model reached an accuracy of 0.83, exceeding standalone textual Capsule Network performance by 0.07 points and other deep learning baselines by up to 0.19 points.
Key insights
- Capsule Networks, specifically the Bi-GRUCapsule, demonstrate high efficacy in extracting nuanced textual features from Persian tweets for suicide stressor detection.
- The integration of auxiliary features like sentiment, polarity, and user engagement metrics with textual data through multimodal fusion substantially boosts classification performance.
- Deep learning architectures offer a scalable and precise solution for identifying psychiatric stressors related to suicide in less-resourced languages like Persian, where manual methods are infeasible.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.