Modeling driver behaviour and the interplay of demographic and environmental factors in traffic violations: A bayesian network approach

Road traffic accidents remain a leading cause of mortality worldwide, with over 1.3 million preventable deaths annually and millions more injured (WHO). Traditional research approaches often fail to capture the complex interplay between demographic, behavioral, and environmental factors contributing to crash outcomes. This study addresses these gaps by employing a Bayesian Network (BN) model to analyze real-world data from Tempe, Arizona, spanning a two-year period. The model integrates variables such as behavioral factors, demographic attributes, and environmental conditions to assess their compound impact on crash severity. The BN model allows for both predictive analyses and causal inferences, providing actionable insights for targeted interventions. Sensitivity analyses highlight key contributors to crash outcomes, with violations like speeding and impaired driving identified as critical behavioral risks, often exacerbated by environmental conditions like low visibility. The study also provides a granular analysis with age-specific risk patterns, emphasizing the need for tailored safety strategies for young, adult, and senior drivers. This research demonstrates the use of Bayesian Networks in capturing risk interactions, offering a comprehensive methodology and a granular driving risk analysis for policymakers and urban planners. By addressing the limitations of traditional approaches, the study contributes to the development of data-driven strategies to mitigate traffic accidents and enhance road safety and propose a novel BN model considering behavioral, demographic and environmental risk factors.