What does this Repository contain ?

Scripts that help in website fingerprinting.

What is TOR ?

The Onion Router TOR is a low latency anonymity network of relays which aims to hide the recipient and the content of communications from a local observer, i.e., an entity that can eavesdrop the traffic between the client and the server. Based on that research, it offers a technology that bounces internet users' and websites' traffic through "relays" run by thousands of volunteers around the world, making it extremely hard for anyone to identify the source of the information or the location of the user. It is extremely useful in totalitarian regimes as it helps users to freely communicate on the web.

What is Website Fingerprinting ?

Website Fingerprinting is an attack on the anonymity of the TOR Network. Initially developed for security purposes, website fingerprinting (also known as device fingerprinting) is a tracking technique capable of identifying individual users based on their browser and device settings. In order for websites to display correctly, your browser makes certain information available about your device, including your screen resolution, operating system, location, and language settings. These details essentially make up the ridges of your digital fingerprint. It helps the attacker to know about the web history of a client and which page the client is browsing through encrypted or anonymization networks.

Workflow

1. It can be concluded that the TOR anonymity network is prone to website fingerprinting, analyse the effect of website fingerprinting.

2. This research focusses on a website fingerprinting technique, K-Fingerprinting, proposed by Hayes that is based on Random Forests and tries to improve the results of the existing state-of-the-art system.

3. The dataset assumes a TOR Hidden Service of 30 websites only. Each Website further contains 100 webpages which are split in 80:20 for the training and the testing set.

4. In the first step previous system is transferred to a model that uses the Random Forest Classifier. In the subsequent steps, work is done on on selecting and computing the features in order to arrive at a conclusion.