Analyzing Interrelationships Between Asphalt Pavement Distresses Across Multiple Lebanese Regions Using Association Rule Mining for Enhanced Maintenance Strategies

Pavement distresses play a pivotal role in evaluating roadway conditions and guiding maintenance strategies. In many instances, these distresses arise from construction deficiencies, substandard material quality, or inadequate maintenance practices, rather than from inherent design shortcomings. Understanding the interrelationships among different types of pavement distress is therefore essential for engineers and decision-makers seeking to enhance pavement performance and prolong service life. This study examines the statistical associations among various asphalt pavement distresses using association rule mining techniques. The dataset was collected from 18 regions in Lebanon, covering a total roadway length of 419.87 km, thereby enabling a comprehensive analysis. Ten primary categories of pavement distress were identified and analyzed using support, confidence, and Lift measures to quantify their co-occurrence patterns and dependency relationships. The results revealed strong statistical associations among most pavement distress types, with particularly strong interactions involving raveling and weathering, longitudinal cracking, alligator cracking, patching, and potholes. Raveling and weathering were the most prevalent distress factors, accounting for 20.88% of total occurrences, whereas block cracking was the least frequent, representing only 0.34%. The joint probability of raveling and weathering occurring with alligator cracking reached 37.38%, while the conditional probability of alligator cracking given the presence of raveling and weathering was 47.65%. Lift analysis further distinguished associations that were stronger than expected based on distress prevalence alone, thereby reducing the influence of frequency-driven relationships. These findings demonstrate the effectiveness of the proposed analytical framework in identifying statistically grounded pavement distress interactions and provide complementary information to support pavement management, maintenance prioritization, and resource allocation.