🛡️ SafeGuard AI - Bullying & Toxicity Detector

A real-time text classification tool built to detect and flag harmful content, including bullying, insults, and toxic comments. This project uses a fine-tuned BERT model to understand context and nuance, promoting safer online interactions.

If you find this project useful or interesting, please consider giving it a ⭐ star!

🚀 Live Demo

You can try the live application here:

https://safegaurd-ai.streamlit.app/

✨ Key Features

• High-Performance Detection: Accurately classifies text into ‘Bullying’ or ‘Not Bullying’ categories.
• State-of-the-Art Model: Powered by a fine-tuned BERT model, allowing it to understand context, sarcasm, and other nuances.
• Interactive UI: A simple and clean user interface built with Streamlit.
• Ready to Use: The repository includes the pre-trained model files, ready for deployment.

📊 Model Performance

The model was fine-tuned on a balanced dataset of over 115,000 text samples, achieving a 93% accuracy. It is particularly effective at identifying harmful content, with a 96% recall on bullying…