PBTR: Pre-training and Bidirectional Semantic Enhanced Trajectory Recovery

Qiming Zhang, Tianxi Liao, Tongyu Zhu, Leilei Sun, Weifeng Lv

Abstract: The advancement of position acquisition technology has enabled the study based on vehicle trajectories. However, limitations in equipment and environmental factors often result in missing track records, significantly impacting the trajectory data quality. It is a fundamental task to restore the missing vehicle tracks within the traffic network structure. Existing research has attempted to address this issue through the construction of neural network models. However, these methods neglect the significance of the bidirectional information of the trajectory and the embedded representation of the trajectory unit. In view of the above problems, we propose a Seq2Seq-based trajectory recovery model that effectively utilizes bidirectional information and generates embedded representations of trajectory units to enhance trajectory recovery performance, which is a Pre-Training and Bidirectional Semantic enhanced Trajectory Recovery model, namely PBTR. Specifically, the road network’s representations extracting time factors are captured by a pre-training technique and a bidirectional semantics encoder is employed to enhance the expressiveness of the model followed by an attentive recurrent network to reconstruct the trajectory. The efficacy of our model is demonstrated through its superior performance on two real-world datasets.