The evaluation of accidents contributing factors by Rough set theory, In Gilan province

Accidents are caused by various parameters and their occurrence cannot be solely related to a single factor. Therefore, various safety studies have been done to represent models for the process of accidents occurrence. One of the new methods to analyze multi variable events and decisions is "Rough Set Theory". By using and taking the advantage of this method, it explores data sets, and by defining independent variables (accident conditions), and dependent variable (the result of accident), the process of accident occurrence can be investigated and the patterns for those accidents that have been frequently repeated under similar conditions, can be extracted. Police-reported crash data for the year 2009 from Gilan province are used in this study. By having this information and using rough set theory, the prevailing patterns for those accidents which occurred under same special conditions were defined. As a result, non-native drivers were involved in damage only accidents and native drivers were involved in injury ones. Damage only accidents were between vehicles, and injury accidents were the result of collision between vehicles and motorcycles or pedestrians. Straight routes were presented in most of the strong patterns. Damage only accidents were between vehicles, and injury accidents were the result of collision between vehicles and motorcycles or pedestrians. Thus, by knowing the repeated and strong patterns, more effective solutions can be adopted.