wsp-lab/adcpg

AdCPG

Code release for the paper entitled "AdCPG: Classifying JavaScript Code Property Graphs with Explanations for Ad and Tracker Blocking", published at CCS 2023.

Requirements

1. Install dependencies.

$ pip install adblockparser
$ pip install escodegen
$ pip install esprima
$ pip install networkx
$ pip install pandas
$ pip install requests
$ pip install scikit-learn
$ pip install selenium-wire
$ pip install torch
$ pip install torch_geometric

2. Install Joern by following the instructions in this link.

Execution

Dataset

JavaScript files are stored in <AdCPG_directory>/data/scripts.

$ cd <AdCPG_directory>
$ python crawler.py

Phase I: Building CPGs

CPGs are stored in <AdCPG_directory>/data/scripts.

$ cd <AdCPG_directory>
$ python builder.py <Joern_directory>

Phase II: Classifying CPGs

Classification results are stored in <AdCPG_directory>/data/results.

$ cd <AdCPG_directory>
$ python classifier.py

Phase III: Generating Explanations

Explanations are stored in <AdCPG_directory>/data/results/explanations.

$ cd <AdCPG_directory>
$ python explainer.py

Citation

@inproceedings{10.1145/3576915.3623084,
    author = {Lee, Changmin and Son, Sooel},
    title = {AdCPG: Classifying JavaScript Code Property Graphs with Explanations for Ad and Tracker Blocking},
    year = {2023},
    booktitle = {Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security},
    pages = {3505–3518},
}