Mining Pattern from Road Accident Data: Role of Road User’s Behaviour and Implications for improving road safety

At the heart of any strategic effort to address a nationwide problem there is data or information. This research tries to view accident data collection and analysis as a system that requires a special view towards understanding the whole and making sense out of it for improved decision making in the effort of reducing the problem of road safety. As part of an information architecture research for road safety data/information management in developing countries, the objective of this machine learning experimental research is to explore and predict the role of road users on possible injury risks. The research employed Classification and Adaptive Regression Trees (CART), TreeNet and RandomForest approaches. To identify relevant patterns and illustrate the performance of the techniques for the road safety domain, road accident data collected from Addis Ababa Traffic Office is used. After collecting the data and format it in the way suitable for the tool used model building and evaluation through prediction success and error rate were the major tasks. Interpretation of the result and recommendation was also made. Empirical results showed that the models could classify accidents with promising accuracy.