Anomalies in Tor usage documentation
Anomalies in Tor usage
Python package that implements the algorighms described in the paper On Identifying Anomalies in Tor Usage with Applications in Detecting Internet Censorship.
There're three date intervals involved:
- the date interval for the "training" data (all that is available by default)
- the date interval to include in the graph (180 days by default)
- the date interval in which the user would like to find anomalies (one day by default)
Author: juga at torproject.org
Copyright (c) 2025-2026, The Tor Project, Inc.
User guide
git clone https://gitlab.torproject.org/tpo/network-health/tor_anomalies
pip install virtualenv
virtualenv ~/.virtualenvs/anomalies
source ~/.virtualenvs/anomalies/bin/activate
pip install -e .
tor_anomalies -h
More documentation at docs/ or
https://tpo.pages.torproject.net/network-health/tor_anomalies/index.html.