Digital Behavioral Biometrics and Privacy: Methods for Improving Business Processes without Compromising Customer Privacy

Abstract

To enable people to interact with online websites and systems, browsers capture a variety of events that occur on the page - such as how a person is moving the computer mouse, what a person clicks on, what a person types, and whether a person is scrolling. These events represent a user's behavior on a page, referred to as the DOM or Document Object Model, and are recorded at a millisecond precision rate (e.g., the exact millisecond timestamp when a key goes down and when it comes back up). Research and practice alike have found that these behavioral events can provide powerful insight into the users' experience, such as whether users are frustrated, and even help distinguish between legitimate and fraudulent users. In this paper, we present six best practices for responsibly collecting these digital behavioral biometric data to help protect user privacy as well as encourage proper interpretation. For each principle, we discuss its rationale and practical application.