Keystroke Dynamics Data Collection and Analyzing from Android Users

Abstract

One of the most widely used behavioural biometrics for second factor authentication in a lot of web services and applications nowadays is keystroke dynamics. Resettable biometrics satisfy a primary usability requirement for authentication systems, which is one of the factors contributing to its immense popularity. Developers and researchers used a variety of machine learning algorithms to identify smartphone users based on their keystroke dynamics, taking advantage of the latest developments in mobile technologies. The ability to gather smartphone user datasets that could be used to train machine learning algorithms and, ultimately, produce accurate predictive models, is the largest challenge facing researchers in this field. In this paper, a native Android application called iProfile is presented. It gathers keystroke dynamics from Android smartphone users. With the help of this application, researchers can now find volunteers to help collect data on keystroke dynamics from anywhere in the world. Researchers can examine how various factors, including hardware brands, users' geolocation, native language text direction, and other factors, affect machine learning classifier accuracy by using our iProfile application. It also contributes to the upkeep of a common benchmark for keystroke dynamics. Having a common benchmark makes it easier for researchers to assess their own work using standardised metrics and data collection techniques. The primary components of the iProfile application, the implementation algorithms, and the database communication protocol are all explained in this paper.