trellixvulnteam/harpo-artifacts_ljkk

Harpo-Artifacts (Ongoing)

Overview

Harpo is a principled learning-based approach to subvert online behavioral advertising. It uses reinforcement learning to adaptively interleave real page visits with fake pages to distort a tracker’s view of a user’s browsing profile.

Accepted paper

Jiang Zhang, Konstantinos Psounis, Muhammad Haroon, Zubair Shafiq. HARPO: Learning to Subvert Online Behavioral Advertising [C]. NDSS, 2022.

Crawler

1. The build-crawler directory contains our scirpt for building our crawling infrastructure. You can firstly run ./build-openwpm.sh to build a docker image for OpenWPM (https://github.com/openwpm/OpenWPM/tree/firefox-52). Then you can run ./build-harpo.sh to build a docker image for running crawling experiments in our Harpo paper. The main function we use for crawling is run_ml_crawl.py, which defines the workflow of crawling experiments. Note that config.json defines the configuration of our browser used during experiments.
2. The run-crawler directory contains our script for running crawling experiments. The main class we use is called docker_api in docer_api.py. It will launch a docker container with openwpm-harpo image and run command sudo /usr/bin/python3 /opt/OpenWPM/run_ml_crawl.py --cfg '{}' in the container.