Analysis on Traffic Accident Severity Considering Surrounding Traffic Characteristics

Abstract

Today, various Big Data is available. But there are few analysis by using plural data source. In Japan, ETC has been populated in our society now. Since ETC records almost all the individual trip on highway. In this study, the new traffic accident factors were investigated by making the characteristic indexes from ETC Big Data. Specifically, the variables expressing peripheral traffic conditions in case of traffic accident in a highway. And a factor that affect traffic accident severity was investigated from candidates for explanatory variables by using ordered probit model. The traffic accident severity was defined independently. As a result of analysis, the possibility that the neighbor car’s characteristics in case of traffic accident such as frequency of use, travel time and monthly travel time affect traffic accident severity was clarified. From newly gained knowledge, the time and point where has potentially hazardous of a serious accident would be clarified.