Multi-Factor Traffic Pattern Discovery Using Cluster Based Association Rule Mining For Smart Traffic Management

Abstract

Efficient traffic management in smart cities requires the discovery of hidden patterns from heterogeneous traffic data generated by modern transportation systems. Traffic datasets often consist of multiple dynamic factors such as traffic volume, vehicle speed, vehicle type distribution, weather conditions, environmental parameters, accident occurrence, and signal status, making traffic pattern analysis a challenging task. To address this issue, this paper proposes a novel hybrid framework named Traffic Pattern Discovery (TPD) using Fuzzy Association Rule Mining (TPD-HFA) for intelligent traffic management. The proposed framework integrates Fuzzy C-Means (FCM) clustering and the Apriori algorithm to identify meaningful traffic behavior patterns from multi-factor traffic data. Initially, Fuzzy C-Means clustering is employed to group traffic records into similar behavioral clusters by handling uncertainty and overlapping traffic conditions more effectively than hard clustering approaches. This clustering process enhances the homogeneity of traffic data and improves the quality of pattern extraction. Subsequently, the Apriori-based Association Rule Mining technique is applied within each cluster to discover frequent traffic patterns and hidden relationships among traffic, environmental, and accident-related attributes. The extracted rules provide valuable insights into congestion formation, accident-prone situations, weather-induced traffic variations, and signal inefficiencies. Experimental analysis demonstrates that the proposed TPD-HFA framework effectively uncovers high-confidence and high-support traffic rules, enabling better decision-making for adaptive traffic signal control, congestion reduction, and road safety enhancement.