Detecting browser fingerprint evolution for identifying unique users

摘要

Unique user identification is a key task within the web analytics data collection process, useful for measuring advertising campaigns effectiveness. The fingerprinting technique consists in tracking user activity on a set of sites by capturing technical information about the browser and the machine that the user uses to navigate the web. Browser fingerprinting has been demonstrated to be an effective method for unique user identification when the device used to navigate the web does not support cookies. However as the attributes used for generating browser fingerprint evolve, multiple distinct fingerprint records are created for the same user, leading to incorrect unique user identification. This paper presents a technique for unique user identification that detects browser fingerprint evolution.