Understanding crowd behavior has attracted tremendous attention from researchers over the years. In this work, we propose an unsupervised approach for crowd scene modeling and anomaly detection using association rules mining. Using object tracklets, we identify events occurring in the scene, demonstrated by the paths or routes objects take while traversing the scene. Allen's interval-based temporal logic is used to extract frequent temporal patterns from the scene. Temporal association rules are generated from these frequent temporal patterns. Our goal is to understand the scene grammar, which is encoded in both the spatial and spatio-temporal patterns. We perform anomaly detection and test the method on a well-known public data.