Factors influencing the patterns of wrong-way driving crashes on freeway exit ramps and median crossovers: Exploration using ‘Eclat’ association rules to promote safety
DOI: 10.1016/j.ijtst.2018.02.001
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study addresses the persistent traffic safety issue of wrong-way driving (WWD) crashes, which frequently result in fatalities or severe injuries. While previous research relied on descriptive statistics or parametric models like logistic regression, these methods often fail to capture complex, non-linear interactions between crash factors or require strict distributional assumptions. To overcome these limitations, the authors employed data mining techniques, specifically the ‘Eclat’ association rule algorithm, to identify significant patterns and interdependencies among factors contributing to WWD crashes. The study broadened the scope of analysis beyond controlled-access highways to include median crossover encroachments on low-speed roadways, aiming to provide comprehensive insights for safety strategies. The researchers analyzed five years (2010–2014) of Louisiana traffic crash data, compiling a dataset of 1,419 WWD crashes involving 2,651 individuals. The data merged information from crash, roadway inventory, and vehicle databases, resulting in 16 variables with 99 categories. Variable importance was initially assessed using a random forest algorithm. The core analysis utilized the Eclat algorithm, a vertical data mining technique efficient for identifying long patterns, combined with the Apriori algorithm for two-item sets. The study optimized support and confidence thresholds using convex optimization to balance rule quantity and interpretability. Association rules were evaluated based on support, confidence, and lift, with lift values greater than 1.0 indicating positive interdependence between antecedent and consequent factors. The results revealed strong associations between specific crash characteristics. The most significant finding was the strong interdependence between fatal crashes and head-on collisions, with a lift value of 3.89, indicating that fatal WWD crashes were nearly four times more likely to be head-on collisions than the general dataset average. Driver impairment emerged as a critical factor, appearing in the top rules across multiple itemsets. For instance, the combination of male drivers and driver condition violations strongly predicted impairment (lift of 4.32). Other significant patterns included the prevalence of crashes in business localities, during off-peak hours, and on two-way roadways without physical separation. The analysis also highlighted that pavement markings, such as yellow lines, were insufficient in preventing median crossover crashes, suggesting a need for physical barriers. Additionally, improper passing violations were strongly associated with daylight conditions, potentially due to driver complacency. The significance of this study lies in its application of machine learning to uncover complex, multi-factor patterns in WWD crashes that traditional statistical methods might miss. By identifying specific combinations of factors—such as the link between male drivers, impairment, and fatal head-on collisions—the findings offer actionable intelligence for state departments of transportation and local agencies. The results support the development of targeted engineering solutions, such as installing physical medians on high-risk two-way roads, and inform safety strategies aimed at reducing driver impairment and addressing specific behavioral patterns during off-peak hours. This approach provides a robust framework for mitigating WWD crashes through evidence-based countermeasures.
Key finding
Fatal wrong-way driving crashes are strongly associated with head-on collisions, male drivers, off-peak hours, and driver impairment, with two-way roadways lacking physical separation being a dominant contextual factor.
Methodology
dataset
Sample size: 1419
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. Discovered via author_sweep_intake on 2026-05-27.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-27 |
| archive | success | canonical_url | — | — | 5 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | semantic_scholar | — | — | 2 | 2026-06-04 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- pre crash contributing factors
- causation analyses
- naturalistic crash near crash
- incidence prevalence
- rail grade crossings
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
- Methodological Resource: dataset resource