Predicting User Acceptance of Automated Vehicles Based on Basic User Characteristics through Big Data Analysis

User acceptance is the key to the adoption of automated vehicles, yet previous models rely heavily on using psychological constructs which are time-consuming to measure to make the prediction. Thus, we establish a simple model to predict user acceptance through big data analysis based on basic user characteristics. According to 456 responses from a survey, the decision tree algorithm was used to build a classification model of users' acceptance of automated vehicles using age, income, gender, highest education, and marital status. The classification model with a maximum depth of four showed a portrait of people who had a high level of acceptance. This portrait was applied to the rest of the participants with 86.75% of the hit rate. We also used this model to select a group of customers to participate in an offline activity to experience real automated vehicles. From comparing the results with another group chosen by random selection, the results showed that in the model-selected group, more customers agreed to participate in the whole activity. We discussed the potential to use the model in customer portrait and company marketing activities.