Santiago Jose de Leon Martinez

Position: Eye Tracking and Recommender Systems

Supervising team: Mária Bieliková (KInIT), Róbert Móro (KInIT)

As part of the MSCA Doctoral Network Eyes4ICU, this topic is focused on eye tracking and how it is applied to recommender systems. While RS have been shown to be successful in their implementation over the past years, there are problems that have come to light, such as model biases and filter bubbles. Moreover, the treatment of implicit feedback requires assumptions that may be reasonable or work in current systems, but should be validated in a scientific manner. Eye tracking data provides a window into how users are processing and interacting with RS during use. It is an additional source of implicit feedback that is much richer and interpretable. Through the use of computational gaze models combined with other feedback, researchers can better understand what information users are receiving throughout their system interaction. A better understanding of users and their states while using systems allows one to begin to better answer the problems of biases, filter bubbles, and validity of current implicit feedback handling.