Classifying user activities using eye fixation clustering, fixation features, and regions of interest

Abstract

A computing device classifies user activities. The device receives eye tracking data for a person viewing a page having multiple contiguous regions. The eye tracking data comprises a temporal sequence of fixations, where each fixation has a duration and a location. The device partitions the fixations into clusters, where each cluster has a consecutive sub-sequence of the fixations. The device assigns a provisional user activity label to each fixation based on a set of characteristics of the fixation. The device also groups together consecutive fixations that have the same label to partition the fixations into groups. For each group that matches a respective cluster, the device retains the provisional label assignment as a final user activity label assigned to each of the fixations in the respective group. The device also reconciles non-matching groups with non-matching clusters, using the regions, to form a set of non-overlapping modified groups.