Intelligent Assessment of Design Layout Based on Eye-Tracking Data Analysis

Numerous design competitions brought new design talents and excellent innovative ideas to the fore. The major request of the review system application is to reduce the burden of manual assessment work. In this study, we attempted to explore the relationship between the indicators of eye movement and the assessment of design works, aiming at developing intelligent methods for evaluating the perception of design works as a primary screening method or providing the suggestion for ranking results. The features of eye-tracking data were extracted to form the fusion dataset. This study applied comprehensive evaluation methods to determine the weights of five kinds of eye movement indicators and obtained the rank values of different design works as the basis of a decision. We compared the experts rating through Kansei Engineering approach and the automatic design assessment ranking result based on eye-tracking data modelling. The results proved that the top 20 design works could be successfully ranked within the top 40, which indicated that the computer evaluation results were concordant with the experts' assessment results to a certain extent. The findings suggest that it is promising in evaluating design works through eye-tracking data analysis, and it will improve design assessment efficiency dramatically.