Phishing Detection
ML-Based URL Classifier
What It Solves
Detecting phishing websites using machine learning-based URL classification. Extracts 200+ features from URLs including entropy analysis, path depth, subdomain detection, and Unicode patterns. Uses ensemble of Random Forest and Gradient Boosting for high accuracy detection with low false positives.
Comparison: Base Paper vs This Project
| Aspect | Base Paper | This Project |
|---|---|---|
| Method | Text-only features | ML-based URL classifier |
| Model | Traditional heuristics | Random Forest + Gradient Boosting ensemble |
| Features | Static checks only | 200+ extracted features (entropy, URL depth, path analysis) |
| Accuracy | ~85% | 99.6% |
| F1-Score | ~0.85 | 0.997 |
| False Positives | High (15%+) | Low (0.34%) |
| Interface | CLI | Professional Streamlit + API |
Tech Stack
Code Structure
phishing-guard/
├── app.py # Main application
├── setup_env.py # Environment setup
├── train.py # Training script
├── email_scanner.py # Email analysis
├── pyproject.toml # Project config
│
├── 03_training/ # Training pipeline
│ ├── dataset.py # Dataset loading
│ ├── train_with_mlflow.py # MLflow integration
│ └── model_manager.py # Model management
│
├── 02_models/ # Trained models
│ ├── phishing_classifier.joblib
│ ├── feature_columns.joblib
│ ├── feature_scaler.joblib
│ ├── classifier.py
│ └── model_metrics.json # 99.6% accuracy
│
├── tests/ # Test suite
│ ├── test_security.py
│ ├── test_comprehensive.py
│ └── test_cases.txt
│
├── docs/ # Documentation
│ ├── architecture.md
│ ├── development-guide.md
│ ├── FINAL_REPORT.md
│ └── TESTING_GUIDE.md
│
├── diagram_prompts/ # UML diagrams
│ ├── 01_system_architecture.md
│ ├── 02_use_case_diagram.md
│ ├── 03_activity_diagram.md
│ └── ...
│
├── release/ # Built executables
│ ├── phishing-guard-desktop
│ └── phishing-guard.sh
│
└── viva/ # Presentation
├── Phishing_Guard_Complete_Documentation.pdf
└── Phishing_Guard_Professional_Presentation.pptx What's Included
- Complete source code (Python/scikit-learn)
- Trained model weights (.joblib)
- Feature extraction pipeline
- One documentation type FREE (IEEE Report / PPT / README)
- Additional documentation: +₹500 per type
- Base paper + citation
- Streamlit + FastAPI interface
- Unit tests + test cases
- Demo videos
- Email support (7 days)
Base Paper
Get This Project
₹5,000 per project
Step 1: Share Your Requirements
Fill this form with your requirements, then message me on WhatsApp:
Fill Requirements FormStep 2: Message Me on WhatsApp
WhatsApp MeStep 3: Pay via UPI
Transfer ₹5,000 to:
8106968013@ibl
After payment, message me the UTR/RRN on WhatsApp
What's Included in Documentation
Each project comes with comprehensive documentation to help you understand and present your project.
Project Report
IEEE format documentation
Presentation Slides
PPT for viva presentation
README Guide
Step-by-step setup instructions
Base Paper
Original research paper cited
Note: These are the documents included. Click "View Details" on each project page to see actual project-specific documentation.