The Application of Multi-Criteria Decision-Making Analysis to Identify Hot Spots: Data Envelopment Analysis and Concordance Analysis

Detection and ranking accident-prone locations or black spots in a transportation network is a basic step in the process of traffic safety improvement. The current study uses two different methods; data envelopment analysis and concordance analysis, as alternative ways for detecting and ranking black spots. The methods are branches of multi-criteria decision making analysis. One of the important characteristics of these methods is their capability of using different factors affecting accidents, considering crashes’ intensities, in ranking hotspots in the network. The current paper is devoted to study hotspots in Qazvin based on two aforesaid methods. These locations are considered to be decision-making units in data envelopment analysis in which inputs and outputs are effective factors on accidents, and different kinds of accidents, respectively. On the other hand, concordance analysis by using a combination of available variables and defining concordance and discordance sets and indices, ranks the locations of interest according to their risks. One of the advantages of these two methods is their multi-criteria nature, which appear to conform better to the pattern of hotspot identification, than the traditional methods. The other advantage (particularly for data envelopment analysis) is to devote attention to the systems’ inputs as well as the concept of efficiency, what is not considered in the conventional methods of hot spot identification. The study on two methods showed that the accuracy of proposed methods is higher due to its tendency to the effective factors on accidents, although having similar results from the two methods, because of different approaches, is not expectable.