In order to build and optimize the official website of low-cost airlines, a user experience analysis and prediction method based on eye movement and facial expression (AU) is proposed. Firstly, six representative official websites of low-cost airlines were selected to carry out user experience experiments. The eye movements and facial expressions in the process were collected. Then, the eye movement behavior and emotional in the process of official website experience were analyzed and compared with the subjective evaluation of user experience. Finally, the correlation analysis was used to extract eye movement and facial expression indicators significantly related to the user experience. Based on the feature vectors composed of these indexes, the classification prediction model of user experience was constructed based on the support vector machine method. The experimental results show that the web design of airlines should balance text content, functional area layout, color and operability in order to achieve good user experience. The user experience prediction model based on “eye movement + AU” can support the real-time user experience monitoring and dynamic evaluation.