Discovering Spatial Co-Clustering Patterns in Collision Data

Abstract

Identifying spatial patterns of collisions is critical for improving the efficiency and effectiveness of traffic enforcement deployment and road safety. Recently, many studies have centred on finding locations with high collision concentration, so-called hotspots. However, most of them only focus on the location information of the collision data, without integrating the non-spatial attributes into analysis. Taking non-spatial attributes into account opens opportunities to reveal attribute-related hotspots that otherwise goes undetected, and can add valuable indicators for explaining those hotspots. In this thesis, we address this problem. We propose a method for identifying the sets of non-spatial attribute-value pairs (AVPs) that together contribute significantly to the spatial clustering of the corresponding collisions. We call such AVP sets Spatial Co-Clustering Patterns (SCCPs). By applying our method on Edmonton’s collision data, we discovered larger numbers of meaningful hotspot patterns than traditional methods did, and revealed the relevant non-spatial indicators for explaining those hotspots.