A DATA MINING APPROACH TO THE IRANIAN LORRY DRIVERS WITH VARIOUS CHARACTERISTIC DRIVER BEHAVIORS, VEHICLES, AND TRAVEL

Transportation and logistics play an important role in the economy of countries, while road transport is one of the most important modes of transportation for load transportation, especially in developing countries. Accidents are directly related to driving offenses, and drivers who commit more offenses are more prone to accidents. Therefore, reducing driving offenses can reduce accidents. Hence, the recognition of common driving offenses among heavy vehicle (truck) drivers and the effective factors in directing them to reduce driving offenses can consequently reduce the frequency and severity of accidents. Thus, there is a necessity for conducting further studies to carry out research in this regard more than ever before. The main objective of this study is to identify and evaluate important factors affecting lorry drivers committing traffic offenses. To this end, the required information was divided into six categories: traffic tonnage, not wearing a seatbelt, unauthorized speed, impaired driving, talking on the phone, lacking a scalp leaf, and these factors are known as dependent variables. Also, the influencing factors in the group of driver characteristics, vehicle, and mileage were considered by using a demographic questionnaire and Driving Behavior Questionnaire (DBQ) and interviews with 420 drivers over 60 days at the Shahid Kheibari Terminal in Mashhad. After correcting or removing incomplete questionnaires, the information of 351 drivers was used for statistical analysis. Besides, statistical analysis of data using the multivariate logistic regression model showed that drivers with discharging and loading of five or six times per month are less likely to commit excess tonnage than drivers with discharging and loading of more than 12 times per month. The results also proved that drivers with less slip behavior and not so-called distraction would be less likely to commit unauthorized speed offenses and 85.4% less likely to commit this violation. Finally, analysis of statistical analysis showed that drivers with aggressive driving behavior were more likely to commit lacking a scalp leaf offense.