🛡️ Phishing Detection Model (FastText)

A lightweight FastText-based model to classify domain names as phishing or clean. It uses supervised learning with wordNgrams=2 for better n-gram feature coverage.

🚀 Installation

Option 1: Build FastText from Source (recommended)

git clone https://github.com/facebookresearch/fastText.git
cd fastText
mkdir build && cd build
cmake ..
make

Option 2: Install via pip (limited support)

pip install fasttext

⚠️ Note: The pip version does not support all FastText features. Compiling from source is recommended.

📦 Usage

Predict a single domain

echo "carreeffoursa.site" | ./fasttext predict phishing_model.bin -

API with Docker

You can also run the model via Docker API:

# Run with Docker Hub image
docker run -p 8080:8080 mstfknn/phishing-fasttext:latest

# Or run with GHCR image
docker run -p 8080:8080 ghcr.io/mstfknn/phishing-fasttext:latest

Example request:

curl -X POST http://localhost:8080/predict \
  -H "Content-Type: application/json" \
  -d '{"domain": "carreeffoursa.site"}'

Example response:

{
  "domain": "carreeffoursa.site",
  "label": "__label__phishing",
  "probability": 0.9734
}

🏋️ Training Info

• Framework: FastText
• Labels: __label__phishing, __label__clean
• Epochs: 10
• Learning rate: 0.5
• wordNgrams: 2

📊 Dataset

The model was trained on mstfknn/phishing-domain-list-2m-plus, a dataset with 2,000,000 domain names labeled as either phishing or clean.

📄 License

This project is licensed under the MIT License.

🔗 Links

• 🤗 Hugging Face Model
• 🐳 Docker Hub Image
• 📦 GitHub Container Registry (GHCR) Image

Star History