Feature Weighted Linguistics Classifier for Predicting Learning Difficulty Using Eye Tracking

Abstract

This article presents a new approach to predict learning difficulty in applications such as e-learning using eye movement and pupil response. We have developed 12 eye response features based on psycholinguistics, contextual information processing, anticipatory behavior analysis, recurrence fixation analysis, and pupillary response. A key aspect of the proposed approach is the temporal analysis of the feature response to the same concept. Results show that variations in eye response to the same concept over time are indicative of learning difficulty. A Feature Weighted Linguistics Classifier (FWLC) was developed to predict learning difficulty in real time. The proposed approach predicts learning difficulty with an accuracy of 90%.