Analytical Methods and Determinants of Frequency and Severity of Road Accidents: A 20-Year Systematic Literature Review

In this systematic literature review (SLR), we use a series of quantitative bibliometric analyses to (1) identify the main papers, journals, and authors publications that make statistical analysis (SA) machine learning (ML) tools as well technological elements smart cities (TESC) Geographic Information Systems predict road traffic accidents (RTAs); (2) determine extent which identified methods are used for RTAs current trends regarding their use; (3) establish relationship between set variables analyzed frequency severity RTAs; (4) gaps in method highlight potential areas future research. A total 3888 papers published January 2000 June 2021, distributed four clusters—RTA + HA SA (SA, n = 399); RTA ML (ML, 858); SC (TESC, 2327); GIS (GIS, 304)—were analyzed. We Accident Analysis Prevention most important journal, Fred Mannering author, The Statistical Crash-Frequency Data: Review Assessment Methodological Alternatives cited publication. Although negative binomial regression was several years, noticed other models based on deep learning, convolutional neural networks, transfer 5G technology, Internet Things, intelligent transport systems have recently emerged suitable alternatives analysis. By introducing new approach computational algorithms data visualization, SLR fills gap area provides clear picture scientific production field. This information is crucial projecting further research developing visualization oriented automation predictions systems.