Discovery of Factors in Road InfrastructuresAffecting Accidents by Data Mining Method

摘要

Recently very ambitious goals to reduce traffic accidents have be arranged by regulating Japanese andEuropean Union governments. To achieve such goals, it is important to extract critical factors causingaccidents from compounded factors. However, it is difficult to decompose such factors into individualorthogonal elements. Therefore, in order to deal with such compounded factors, we look forapplication of data mining method. To apply data mining method to amounts of data, we mustunderstand the feature of data. We partly analyzed the data using data mining method, and not onlyunderstood the feature but also discovered trends of data leading to extraction of accident factors. Inthis paper, we report this process and discuss future prospects.