We have developed a new driving assistant system that can help low-skilled drivers improve their driving skill. There are three key contributions. First, we have developed a statistical method to extract distinctions between high- and low-skilled drivers based on AdaBoost, which selects a small number of critical operation features between high- and low-skilled drivers. The second is that a driving skill evaluation model is built based on extracted features. Finally, we performed a series of experiments using a driving simulator, in which advice based on extracted features was supplied to low skilled drivers. Result shows advices were effective improved drivers' driving skill.