Analysis of Road Accidents through Multiple Mining Techniques

ABSTRACT Road accident is unpredictable and uncertain in entire world; India is the highest one among in the world. Road accident is one of main cause of property damage, unnatural deaths and disability, Road accident data analysis leads to improve the safety. In this paper we proposing two mining techniques decision tree and IF THEN rule to analyze the accident data. For this experiment we have 1,040,051 accident records from the year 2009 to 2015. Each record contains 30 factors; we have selected four major factors (Road type, Speed, Day of week and area). The decision tree analysis shows that the number of road accident occurred on Single carriageway is more (Average 2.69 %), road accident occurred in one-way street, Sunday & Saturday is less. By using this analysis can improve efficiently the level of Road accident safety management. Keywords: Road accident; Decision tree, IF THEN; Analysis; Safety Management