Skip to content

README.md

Latest commit

 

History

History
124 lines (95 loc) · 4.38 KB

README.md

File metadata and controls

124 lines (95 loc) · 4.38 KB

Keystroke-Dynamics-DataGen

A dataset generator for keystroke dynamics research—a biometric technique to distinguish users based on their typing patterns.

This tool records keystroke timing features and generates JSON-formatted datasets, which can be used for user authentication, security applications, and behavioral biometrics research.

📌 Features Recorded

This script captures three key typing characteristics:

✔ Hold Time – Duration between key press (keydown) and key release (keyup).
✔ Keydown-Keydown Time – Time between pressing two consecutive keys (keydown → keydown).
✔ Keyup-Keydown Time – Time between releasing one key and pressing the next (keyup → keydown).

👉 Refer to this research paper 📄 Keystroke Dynamics by CMU for dataset format.

🚀 Getting Started

1️⃣ Clone the Repository

git clone https://github.com/nileshprasad137/keystroke-dynamics-datagen.git
cd keystroke-dynamics-datagen

2️⃣ Set Up the Environment

Create a virtual environment and install dependencies.

python3 -m venv .venv
source .venv/bin/activate  # For Linux & macOS
# On Windows, use: .venv\Scripts�ctivate
pip install -r requirements.txt

3️⃣ Prepare the Dataset

  • Place the original dataset (DSL-StrongPasswordData.csv) in the root folder.
  • Copy the dataset into the edited_dataset/ folder for appending new data without modifying the original file.

🎬 Running the Keystroke Recorder

To record keystrokes, run the script:

python record_keystroke.py

🔹 The script will prompt you to enter a name for the user recording the keystrokes, e.g., "Nilesh_rep1".
🔹 Modify frequency_password_entry inside record_keystroke.py to set how many times the user needs to enter the password.
🔹 The script will generate keystroke timing data in JSON format inside the output/ folder.

Example

Enter your name: Nilesh_rep1
Enter '.tie5Roanl': .tie5Roanl
Password correct!
Keystroke data saved successfully!

✅ Output will be stored as:

output/Nilesh_rep1_timings.json

📌 Appending New Keystroke Data to Dataset

Once recordings are complete, run the script below to append the new data to edited_dataset/:

python append_in_dataset.py

✔ If you entered the username "Nilesh_rep1", provide "Nilesh" as input when prompted.
✔ Your recorded keystrokes will be added to the dataset in the edited_dataset/ folder.
✔ Original dataset remains unaffected.

🔹 Supported Platforms

✅ Linux & macOS (Uses pynput instead of pyxhook).
✅ Windows Compatibility → Future support using keyboard module.

🛠️ Modifications & Customizations

You can modify the following parameters:

Parameter Description
frequency_password_entry Number of times the user must enter the password.
DEFAULT_PASSWORD Change the predefined password to match your study.
record_keystroke.py Modify the script to capture additional typing features.

📢 Contributions & Improvements

🚀 Looking for contributors!
We welcome contributions to:
✔ Improve cross-platform compatibility (Windows support).
✔ Enhance data visualization tools for keystroke analysis.
✔ Add ML models for authentication & anomaly detection.

To contribute:

  1. Fork the repository.
  2. Create a feature branch (git checkout -b feature-new).
  3. Commit changes (git commit -m "Added Windows support").
  4. Push the branch (git push origin feature-new).
  5. Submit a pull request.

🙌 Contributions are highly appreciated!

📜 License

This project is licensed under the MIT License. Feel free to use, modify, and distribute.

🔗 Related Work

  • 📄 Research Paper on Keystroke Dynamics → CMU Keystroke Study
  • 🔍 Example Use Cases → Behavioral biometrics, continuous authentication, fraud detection.
  • 🏆 Potential Applications → Banking security, anti-bot verification, identity protection.

💡 Need Help? 📧 Contact: nileshprasad137@gmail.com
🚀 Star this repo if you found it useful! ⭐