Discovering Insightful Rules among Truck Crash Characteristics using Apriori Algorithm

Hong, Jungyeol; Tamakloe, Reuben; Park, Dongjoo · 2020 · Crossref

DOI: 10.1155/2020/4323816

archive: archived pipeline: cataloged

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the limitations of traditional parametric models in analyzing freight truck crash data, specifically their reliance on predetermined assumptions about relationships between variables. To overcome these constraints, the authors applied Association Rules Mining (ARM), a nonparametric data mining technique, to identify hidden patterns and interrelationships among risk factors contributing to truck-involved crashes on Korean expressways. The research aims to provide transport stakeholders with actionable insights for policy development by quantifying the strength of associations between crash antecedents and consequents without specifying dependent or independent variables. The methodology employed the Apriori algorithm to analyze a dataset of 19,038 truck-involved crashes recorded on South Korean expressways between 2008 and 2017. The data, sourced from the Korean Expressway Corporation, included 17 explanatory items with 98 subitems covering factors such as driver age, weather conditions, vehicle weight, roadway geometry, and crash type. The ARM process involved determining optimal minimum support ($\alpha$) and confidence ($\beta$) thresholds through iterative trials. The algorithm generated 90,951 association rules, which were then evaluated using the lift metric to determine the strength of association between risk factors. Rules with a lift value greater than 1 were identified as strong associations indicating positive interdependence between the antecedent and consequent items. The analysis revealed distinct patterns in crash contributory factors across various segment types and environmental conditions. Key findings indicated that overspeeding with medium-weight trucks was highly associated with crashes occurring during rainy weather. Conversely, drowsy driving was strongly correlated with crashes during fine weather conditions, particularly in the evening. Furthermore, segment-related crashes were primarily linked to driver faults and specific roadway geometry features. The results highlighted that while most crashes occurred on mainline straight roads under fine weather, the specific combination of driver behavior, vehicle characteristics, and environmental conditions significantly influenced crash likelihood. The significance of this research lies in its application of a nonparametric approach to traffic safety, offering a method that avoids the assumption violations common in parametric modeling. By identifying specific, easily understandable causal relationships among interrelated crash factors, the study provides policymakers and researchers with targeted suggestions for reducing freight truck crashes. The findings underscore the importance of considering the interaction between driver behavior (such as drowsiness or speeding) and environmental contexts (weather and time of day) when designing safety interventions, moving beyond isolated factor analysis to a more holistic understanding of crash dynamics.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• pre crash contributing factors

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes