Inferring Document Readability by Integrating Eye Movement Features

Latest research has shown that the readability of documents plays an important role in information seeking and acquisition, especially for non-domain-expert users. Classical document readability measures are based on surface text features, independent of users. In this paper, we propose to integrate text features with the users’ eye movement features. The latter is expected to reflect a user’s reading level, thus can be used to measure document readability in a personalized way. Based on the eye tracking data collected from our preliminary user evaluation, we investigated the impacts of different features on document readability prediction. The results show that the combination of text and eye movement features has a higher correlation with human judgments than using either of them individually.