Understanding User Attention to Adaptive Hints with Eye-Tracking In this talk, I will discuss research in the field of intelligent, user-adaptive interface agents. Specifically, the work is in relation to understanding the effectiveness of user-adaptive hints and how to improve it. User-adaptive hints are personalized hints generated by autonomous interface agents to respond to each userís specific need based on a user model generated in real time from user actions and possibly other tracked behaviors. We will present our test bed, Prime Climb, a user-adaptive educational game that provides individualized support for learning number factorization skills. In Prime Climb, the support is in the form of personalized hints based on a probabilistic model of user learning. It is challenging to provide meaningful and effective user-adaptive hints in an educational game as this support requires a trade off between fostering learning and maintaining game engagement. Previous studies with Prime Climb indicated that players may not always be paying attention to the hints, even when they are justified. In this talk, I will discuss work on using eye tracking data on user attention patterns to better understand if and how players process personalized hints, with the long term goal of making hint delivery more effective.
Visit the LCI Forum page