SAMO: A Sequential Pattern Mining Model for Evaluating Road Criticality in Urban Traffic Networks

This paper proposed a Sequential pAttern Mining mOdel (SAMO) for discovering critical links and evaluating road importance in transportation systems. SAMO introduces novel criticality indices derived from association rule mining and data analysis of vehicular trajectory data. These indices are designed to assess the criticality of road links within a network by leveraging mining frequent patterns and extract meaningful associations from the trajectory data. SAMO prioritizes links that feature prominently in frequent trajectory patterns, with additional weight given to patterns with high confidence levels. We evaluated the performance of our indices within machine learning prediction models alongside with the traditional indices used in the literature. Our results demonstrate that integrating our proposed indices improves prediction accuracy across various models, with the indices consistently ranking among the top features. This indicates the promising potential of our indices for evaluating and assessing critical links in transportation networks.