Assessment of Driver's Comprehension towards Traffic Signs Based on Data Mining Method

In order to achieve the most complex forms of human behavior، especially to evaluate the performance and behavior of drivers when seeing the traffic signs، utilization of statistical methods based on sampling of the studied population، is one of the suitable ways for research in behavioral sciences. In this study، The CART algorithm as a non-parametric model and a sub division of data searching field has been used، where there is no need to user’s interference. In spite of the high applicability of this method in analysis of safety and accidents issues، the researchers have not paid sufficient attention to it، due to its high complexity. Drivers usually react differently in different conditions and circumstances. This study includes analysis of questionnaires distributed among 527 drivers. Results show that factors like the time of driving، the monthly income of drivers and their occupation، are quite important when their paying attention to traffic signs is under question. Factors like age and genre are not very important in seeing and paying attention to the traffic signs by drivers، although these are effective in creating accidents.