Lane-changing behavior significantly affects traffic safety, often leading to severe casualties and substantial property damage. A lane-changing safety warning model can help drivers make safer lane-changing decisions, thereby reducing accidents. This study is based on the US-101 segment of the NGSIM dataset, where sample data of lane-changing passenger vehicles were selected following specific filtering criteria. A set of driving style indicators related to driving safety was chosen. Factor analysis was employed to reduce the dimensionality of the selected indicators, and the k-means clustering method was then used to classify driving styles into two categories: conservative and aggressive. The results show that aggressive drivers exhibit greater speed fluctuations compared to conservative drivers. Therefore, in the development of a lane-changing safety warning model based on Time to Collision (TTC), an additional minimum safe distance was introduced as a secondary warning criterion for aggressive drivers, aiming to reduce false warnings caused by sudden speed changes. Validation results indicated that the model achieved an accuracy of 82.7%, demonstrating its effectiveness. This warning model provides a foundation for further research into the safety of lane-changing behavior on highways and the reduction of lane-changing accidents.