Devices such as smart phones and tablet PCs of various sizes have become increasingly popular, finding new applications, including in-car audio systems. This paper proposes a new car audio system. In the architecture, music data is stored in an online database, which users are then able to select a genre of music or a playlist from through a 2D interface. Self-organizing map, depending on a personalized distance function and music contents, is utilized to map music tracks to the interface. With this data model and interface, drivers can easily select the type of music they want to listen to without interfering with their driving. Artificial neural networks record and analyze user preference, allowing the system to select and order the music tracks to be played automatically. Experiments have shown that the system satisfies user requirements.