Using large-scale electric vehicle charging behavior data in Beijing, this study reveals the charging behavior characteristics of electric vehicle users and their individualized needs through analysis and modeling. First, the data were effectively processed and feature extracted, and a database containing many charging records was established. Then, the Gaussian mixture model algorithm was applied for clustering analysis, and four charging behavior patterns were successfully identified: short-duration low-power charging mode, balanced charging mode, long-duration high-power charging mode, and high-efficiency fast charging mode. Finally, the personalized characteristics of individual users are found by profiling their charging behaviors.