AraEyebility Corpus (Eye-Tracking Data for Arabic Text Readability Corpus)

AraEyebility is the first Arabic eye-tracking dataset designed to capture real-time reading behaviors. The eye movements of 15 native Arabic speakers were recorded as they silently read 587 paragraphs, totaling 57,617 words, spanning diverse topics and linguistic features. This corpus serves as a valuable resource for analyzing a wide range of text and gaze-related features—including fixation, saccades, regressions, and pupil dynamics—which are key indicators of reading effort in Arabic reading research.