Empowering digital well-being with AI-driven insights into mental health risks on social media.
Boop is an innovative mental health platform designed to identify and support individuals showing signs of self-harm and suicidal ideation on social media. Utilizing cutting-edge natural language processing (NLP) through a fine-tuned BERT model, Boop analyzes social media content to detect emotional distress with high accuracy. Integrated with Reddit via APIs, Boop offers a proactive approach to mental health monitoring and intervention.
- AI-Powered Analysis: Leverages a finely-tuned BERT model for advanced sentiment analysis, achieving 94% accuracy in detecting potential self-harm or suicidal thoughts.
- Analysis of Reddit Comments: Utilizes reddit-user-to-sqlite to pull Reddit comment and post data for analysis
- User-Friendly Interface: A Flask-built application paired with a dynamic frontend, ensuring ease of access and interaction for users seeking mental health insights.
- Awarded "Best Project in Mental Hack" for enhancing digital well-being and innovative use of machine learning for mental health support.
- Recognized for its innovative use of AI and API integrations to facilitate immediate support and intervention for individuals exhibiting signs of mental distress.
To set up Boop for development or personal use, follow these steps:
1. clone git repo
git clone https://github.com/nick-ching23/boop.git
cd boop
2. Download the model binary and move it to 'boop' folder
- Google Drive Model Link
- Move model_v2.pth to the folder 'boop'
2. Set Up a Virtual Environment
python3 -m venv venv
source venv/bin/activate # On Unix/macOS
.\venv\Scripts\activate # On Windows
3. Install Dependencies
pip install -r requirements.txt
4. Run Application
python3 main.py
After launching Boop, navigate to http://localhost:5000 in your web browser to access the application. Users can interact with the platform to understand the AI's analysis and receive guidance on potential intervention strategies.
Boop is released under the MIT License. See the LICENSE file for more details.