Cette page appartient aux archives web de l'EPFL et n'est plus tenue à jour.
This page belongs to EPFL's web archive and is no longer updated.

Eye-Tracking for Student Modeling in Exploratory Learning Environments

Eye-Tracking for Student Modeling in Exploratory Learning Environments
Cristina Conati

Department of Computer Science, University of British Columbia, Canada

Department of Information and Communication Technology, University of Trento, Italy

December 14th, 14:00, BC 129


Artificial intelligence has been successfully coupled with cognitive science and educational technology to devise Intelligent Learning Environments that provide computer-based individualized instruction. Providing individualized instruction involves building a model of student traits relevant to adequately tailoring the interaction, i.e., a student model. The relevant student traits may include simple performance measures (such as correctness of interface actions), domain-dependent cognitive traits (such as knowledge and goals) or meta-cognitive reasoning processes that cut across tasks and domains. Arguably, the higher the level of the traits to be captured, the more difficult it is to assess them unobtrusively from simple interaction events. This problem has generated a stream of research on using innovative sensing devices to enrich the information available to a student model.

In this talk, I will contribute to this line of research by presenting results on using on-line eye-tracking information to inform a student model designed to assess student meta-cognitive behavior during interaction with an environment for exploration-based learning. I will first describe the empirical work we did to understand the relevant meta-cognitive behaviors to be modeled. Then, I will illustrate the probabilistic model we designed to capture these behaviors with the help of on-line information on user attention patterns derived from eye-tracking data. I will show that gaze-tracking data can significantly improve the model?s capability to accurately predict student?s meta-cognitive processes and consequent learning, compared to lower level, time-based evidence. Time permitting, I will also discuss work we have done on using pupil-dilation information, also gathered through eye-tracking data, to further improve model accuracy.


Dr. Conati is an Associate Professor of Computer Science at the University of British Columbia, and she is currently spending her sabbatical year in the Department of Information and Communication Technology of the University of Trento, Italy. She received a ?Laurea? degree (M.Sc. equivalent) in Computer Science at the University of Milan, Italy (1988), as well as a M.Sc. (1996) and Ph.D. (1999) in Artificial Intelligence at the University of Pittsburgh. Dr. Conati?s research goal is to integrate research in Artificial Intelligence (AI), Cognitive Science and Human Computer Interaction (HCI) to make complex interactive systems increasingly more effective and adaptive to the users? needs. Her areas of interest include Adaptive Interfaces, Intelligent Tutoring Systems, User Modeling, and Affective Computing. Dr. Conati has served on program committees and as a reviewer for major AI and HCI conferences/journals, and she is program co-chair of User Modeling 2007, the 11th International Conference on User Modeling. She published over 40 strictly referred articles, and her research has received awards from the International Conference on User Modeling, the International Conference of AI in Education and the Journal of User Modeling and User Adapted Interaction.
Posted by Florence Colomb on Friday 1 December 2006 at 17:34