A decision tree approach for traffic accident analysis of saskatchewan highways

Identifying the major contributing factors to traffic collisions and their severity will assist highway safety improvement initiatives by improved facility design and educational program to address the needs due to the changes in demographics. The traffic collision data used in this study has been collected over the last 20 years on the rural highways and urban streets from Saskatchewan, Canada. In order to determine the major factors contributing to traffic collisions and their severity, we present a data mining model using ID3 and C4.5 decision tree algorithms to analyze the traffic collision data. The experiment results from this study will show that the developed data mining model using decision tree can effectively classify the major contributing factors to traffic collisions and their collision severity for different groups of people with good accuracy. The data mining model is evaluated and compared with a commercial software package Weka. Recommendations drawn from the study results for traffic safety improvements are presented.