In this work a method of analysis of descriptive accident features that enables the comparison of road accident patterns has been proposed. The patterns are derived from cross tabulation of qualitative features of road accidents and some ratios are defined that join together pieces of infonI1ation coming from the patterns. The ratios concern two aspects: proportional similarity and structural similarity of road accident patterns. In addition, an algorithm that makes use of the ratios and cluster analysis in identifying accident patterns predominant in the national road network has been worked out. The elaborated method can be applied to detailed as well as to general road accident analyses. In both cases road accident patterns can be examined. In detail (microscopic) analysis, accident patterns for two randomly selected road segments or for a set of road sections or road spots of high accident risk can be compared. In general (microscopic) analysis predominant accident patterns for a road network can be identified and compared. (Author/publisher) For the covering entry of this conference, please see ITRD abstract No. E208120.
Samenvatting