Automatic discovery of browser fingerprinting attributes

Résumé

Users are presented with an ever-increasing number of choices to connect to the Internet. From desktops, laptops, tablets and smartphones, anyone can find the device that best suits their needs while factoring mobility, size or processing power. However, the diversity of modern devices is so great today that it opened the door to a technique called browser fingerprinting. By collecting a set of information related to a user’s device (browser, operating system and hardware), any third-party can build a fingerprint of the device and use it to track an individual online. Over the years, new fingerprinting methods have been found through manual analysis but there is a need for methods that can discover differences between methods automatically.

In this internship, the student will work on extending the work done by Scharwz et al. to discover new methods that could be used for fingerprinting, especially by triggering functions that require parameters. Then, in a second part, the student will evaluate her new methods against a wide variety of web browsers to identify potential privacy-intrusive information.

Mots-clés

browser fingerprinting

Équipe

Spirals

Encadrants

Pierre Laperdrix(https://plaperdr.github.io/), Romain Rouvoy(http://romain.rouvoy.fr/)

Présentation détaillée

Subject

Users are presented with an ever-increasing number of choices to connect to the Internet. From desktops, laptops, tablets and smartphones, anyone can find the device that best suits their needs while factoring mobility, size or processing power. However, the diversity of modern devices is so great today that it opened the door to a technique called browser fingerprinting. By collecting a set of information related to a user’s device (browser, operating system and hardware), any third-party can build a fingerprint of the device and use it to track an individual online. Over the years, new fingerprinting methods have been found through manual analysis but there is a need for methods that can discover differences between methods automatically.

In this internship, the student will work on extending the work done by Scharwz et al. to discover new methods that could be used for fingerprinting, especially by triggering functions that require parameters. Then, in a second part, the student will evaluate her new methods against a wide variety of web browsers to identify potential privacy-intrusive information.

Prerequisites

Experience in JavaScript is optional but strongly recommended.

Bibliography

Comments

For additional information, contact either Pierre Laperdrix or Romain Rouvoy by mail.