Gaze-based interest detection on newspaper articles

Eye tracking measures have been used to recognize cognitive states involving mental workload, comprehension, and self-confidence in the task of reading. In this paper, we present how these measures can be used to detect the interest of a reader. From the reading behavior of 13 university students on 18 newspaper articles, we have extracted features related to fixations, saccades, blinks and pupil diameters to detect which documents each participant finds interesting or uninteresting. We have classified their level of interests into four classes with an accuracy of 44% using eye movements, and it has increased to 62% if a survey about subjective comprehension is included. This research can be incorporated in the real-time prediction of a user's interest while reading, for the betterment of future designs of human-document interaction.