Spatial and Temporal Geovisualisation and Data Mining of Road Traffic Accidents in Christchurch, New Zealand

摘要

This paper outlines the development of a method for using Kernel Estimation cluster analysis techniques to automatically identify road traffic accident 'black spots' and 'black areas'. A Novel data-mining approach has been developed - adding to the generic exploratory spatial analysis toolkit. Christchurch, New Zealand, was selected as the study area and data from the LTNZ crash database was used to trial the technique. A GIS and Python scripting was used to implement the solution, combining spatial data for average traffic flows with the recorded accident locations. Kernel Estimation was able to identify the accident clusters, and when used in conjunction with Monte Carlo simulation techniques, was able to identify statistically significant clusters.