Abstract: The traditional personalized preferences are limited to sensibility analysis, and the user
perceptual preferences of a particular product gotten. The sensibility analysis is subjectivity and
instability, and the personal preferences are lack of scientific theoretical basis. The eye movement
data is put forward a prediction method of users personalized. The multivariate regression equation is
established based on the eye movement data and design elements. The patterns of the batik
personalized user preferences are explored, and the weight of different design elements is gotten to
obtained different solutions according to different user personalization.

Key words: personalization, preferences, eye movement data, multivariate regression