xlin29 / fpclassifier

A Machine Learning-based Approach to Detect Browser Fingerprinting

Description

We propose a machine learning-based approach utilizing VisibleV8(VV8)[1] to accurately detect browser fingerprinting. Our system outperforms the state-of-the-art with an improvement in recall of around 4.8%. We discover JavaScript APIs that particularly appear on fingerprinting scripts from the important features. It shows that our system can be used to perform a large-scale measurement to uncover more previously unreported uses of JavaScript APIs by stateless tracking.

Usage

• visible_v8logs_crawl directory
  • crawl.py: collect script execution.
  • extract_features.py: parse VV8 logs and convert features to a feature matrix.
• logs directory: sample VV8 logs
• MLmodel directory
  • code directory: model evaluation, feature preparation, label construction, and dimension reduction
  • features_v2 directory: data, feature, and labels in different rounds
  • results directory: results and performance evaluation
  • run_logs_err directory: model execution logs
  • saved_models directory: best models

References

[1] Jueckstock, Jordan, and Alexandros Kapravelos. "VisibleV8: In-browser monitoring of JavaScript in the wild." Proceedings of the Internet Measurement Conference. 2019.

fpclassifier's Issues

Hello bro

Hello bro, can you give me a copy of the VisibleV8 browser. I don't use Docker. thanks in advance