AI Phishing Email Detector
Defend. Detect. Automate.
An AI-powered email classification system designed to detect phishing emails based on their content and header structure.
Overview
This project is an AI-powered email classification system designed to detect phishing emails based on their content and header structure. It uses machine learning for classification and features a user-friendly GUI built with Tkinter for real-time analysis.
Project Objectives
- Analyze email content and headers for phishing indicators.
- Train a supervised machine learning model (Naive Bayes / Logistic Regression).
- Build a GUI for easy email input and real-time phishing detection.
- Achieve high accuracy with explainable results.
Key Features
Supervised ML Engine
Classifies emails using trained models (Naive Bayes / Logistic Regression).
Heuristic Extraction
Scans suspicious keywords, header anomalies (e.g. mismatched sender path), URL structures, and attachments.
Tkinter Desktop GUI
Load or paste an email block into a desktop program for instant threat predictions.
Metrics Indicators
Displays classification reasons and indicators to explain why a decision was reached.
Project Structure
AI-phishing-email-detector/
├── phishing_detector.py # Email parsing, feature extraction, model training
├── gui.py # Tkinter GUI for real-time detection
├── model.pkl # Trained ML model file
├── data/ # Sample datasets (PhishTank, SpamAssassin)
├── requirements.txt # Python dependencies
└── README.md # Project documentation
Model Evaluation
Our model evaluates incoming traffic against historical classification benchmarks using standard split training sets:
Future Improvements
- Add support for deep learning models (e.g. LSTM, BERT).
- Deploy as a browser/email plugin for real-time protection.
- Add explanation for classification decisions using SHAP/LIME tools.