Eye-Movement and Touch Dynamics: A Proposed Approach for Activity Recognition of a Web User

Abstract

Behavioral Profiling is the way in which the user interacts with the mobile sensors and services. Identifying the context in which users' interaction occur is an important step toward automatic interpretation of behavior. Activity Recognition and online task monitoring are required for various context aware applications. As a result, two sources of behavioral biometric data are investigated for the development of user Web identification models. The adopted approach is based on eye-gaze movements along with touch dynamics interactions. Fusing different behavioral biometric traits, i.e, the unique characteristics of eye movements and the distinctive way a person touches on a touchscreen device, can improve the identification accuracy.