Identification and Removal of Accident-Prone Locations Using Spatial Data Mining

Road accidents are responsible for a great amount of morbidity and mortality, hence, public health safety on transport channels is a major concern in the present scenario. The present study shows a method for identification and removal of accident-prone locations on roads (hot spots). The proposed methodology includes road network analysis followed by hot spot analysis using Kernel Density Estimation (KDE), buffer operation of Geographic Information System (GIS) and spatial data mining techniques. Various causes for occurrences of accidents are further correlated with identified hot spots. For the present study, accident records on internal roads of an engineering institute have been used. The results reflect that the junction points and low visibility at turns are the main causes behind the occurrences of accidents at identified hot spots. The obtained results can be further used by institute administration for implementing accident prevention measures and removal of hot spots.