Using Eye-Tracking Data for High-Level User Modeling in Adaptive Interfaces

Cristina Conati, Christina Merten, Saleema Amershi, Kasia Muldner

Abstract: In recent years, there has been substantial research on exploring how AI can contribute to Human-Computer Interaction by enabling an interface to understand a user's needs and act accordingly. Understanding user needs is especially challenging when it involves assessing the user's high-level mental states not easily reflected by interface actions. In this paper, we present our results on using eye-tracking data to model such mental states during interaction with adaptive educational software. We then discuss the implications of our research for Intelligent User Interfaces.