Ontology-oriented Method for Road Accident Geoanalysis

Location intelligence is an important approach in the geoanalysis of the urban environment. In this paper, the main attention is paid to the development of ontology-oriented location intelligence method for road accident geoanalysis by means of ontology matching for merging information from heterogeneous sources. The task of accident analysis is to identify potentially dangerous sections of the road network, which can be used to build routes along minimally dangerous roads. The novelty of the method is the integration of heterogeneous sources of spatio-temporal information in a common geoanalysis system by converting them into an ontology and conducting a verification and analysis of the frequency of accidents on individual sections of roads to determine the most frequent type and cause of accidents. The method was tested on traffic accident data from open information source in St. Petersburg, Russia. With the method data about road accidents was verified from various sources and clusters were located with high concentration of accidents.