The design of driver-vehicle interface has to base on the driver`s compatibility for the high operation performance. The BPM (Blind-Pointing Method) model was developed to extract driver`s mental model in the form of cognitive map for the design of compatible in-vehicle interface. However, in this model, established analysis method could not extract the driver`s detailed cognitive map. So, in this research, objective and detailed analysis method was suggested to complement of the incompleteness of existing analysis method in the BPM (Blind-Pointing Method) model. The K-means clustering analysis was used for the new analysis method, and more detailed cognitive map was extracted by suggested analysis method.