GPS Trajectories provide a valuable opportunity for us to understand the human behaviors of road users and detect adverse and/or malignant events. In this chapter, we intend to detect anomalous routes through an in-depth comparison between the target trace and historical “normal” ones. Specifically, a new approach is proposed to detect anomalous trajectory “on-the-fly” and find out which segments cause the anomaly. After that, we conduct a systematical analysis on nearly 43,800 anomalous trajectories derived from 7600 taxis in a month. It reveals that most anomalous trips are caused by deliberate fraudulent decisions by greedy taxi drivers. At last, a group of experiments has been conducted to evaluate the effectiveness of the isolation-based online anomalous trajectory (iBOAT) algorithm. iBOAT algorithm shows superior performance compared with the baseline methods (AUC ≥ 0.99).
iBOAT: Detecting Anomalous Trajectories On-the-Fly
Enabling Smart Urban Services with GPS Trajectory Data ; Chapter : 6 ; 125-146
2021-04-02
22 pages
Article/Chapter (Book)
Electronic Resource
English