With the development of cross-border e-commerce model and the rapid growth of import trade volume, endless methods are being employed to use the cross-border e-commerce retail import channels for committing smuggling crimes. The process of handling relevant data by the customs anti-smuggling department is often very difficult. Based on the intelligence thinking of three-lists matching, the present study investigates the nature of smuggling crimes by using cross-border e-commerce retail import channels, proposes the idea of identifying the nature of cases by matching orders, payment orders and logistics orders, and realizes the risk classification method by combining the principle of three-lists data and RFM model with cluster analysis. Further, by developing a T-R model, the intelligent risk classification, simplification of case nature identification and the visualization of data was realized, so as to improve the text processing ability and intelligence analysis ability of customs anti-smuggling department in investigating and handling relevant cases. Additionally, the results of the study can assist the customs anti-smuggling department to carry out the supervision and anti-smuggling work of cross-border e-commerce retail import.