Spaces:
Sleeping
title: Phishing Email Detector
emoji: 🎣
sdk: docker
app_port: 7860
Phishing Email Detector 🎣
This project is a web-based tool designed to help users identify potentially malicious phishing emails. By pasting the text content of an email, the application leverages a fine-tuned transformer model from the Hugging Face Hub to analyze the content and classify its likelihood of being a phishing attempt.
It serves as a practical, end-to-end example of building and deploying a machine learning application as an interactive web service.
Key Features
Simple Web Interface: An easy-to-use text area for pasting email content for analysis.
Real-Time Analysis: Utilizes a DistilBERT-based model to provide instant classification.
Clear Predictions: Outputs a primary classification (e.g., "Phishing Link Detected", "Legitimate Email") along with a confidence score.
Detailed Breakdown: Displays the model's confidence scores across all possible output labels for greater transparency.
Containerized & Reproducible: Packaged with Docker, ensuring a consistent environment for both development and deployment.
Tech Stack
Backend: Python, Flask, Gunicorn
Machine Learning: Hugging Face Transformers, PyTorch
Frontend: HTML, CSS (via Jinja2 templates)
Deployment: Docker, Hugging Face Spaces
Live Demo
🚀 You can try the live application here: