Introduction to the Special Issue on Intelligent Trajectory Analytics: Part I

We are delighted to present this special issue on Intelligent Trajectory Analytics. Over the past decades, a broad range of techniques have been proposed for processing, managing, and mining trajectory data. It enabled and helped government agencies and businesses to better understand the mobility behavior of their citizens and customers, which is crucial for a variety of applications such as smart city and transportation, public health and safety, environmental management, and location-based services. The purpose of this special issue is to provide a forum for researchers and practitioners in academia and industry to present their latest research findings and engineering experiences in developing cutting-edge techniques for intelligent trajectory data analytics. This special issue consists of two parts. In Part 1, the guest editors selected 10 contributions that cover varying topics within this theme, ranging from car-hailing services to location predictions, from representation learning to trajectory generation. Wang et al. in “PassengerMobility Prediction via Representation Learning for Dynamic Directed and Weighted Graph” propose a novel spatio-temporal graph attention network (Gallat), which can learn a more powerful representation on dynamic directed and weighted graph, for passenger demand prediction. Hu et al. in “Instant Basketball Defensive Trajectory Generation” develop an autoregressive generative model for instantly producing basketball defensive trajectory, with focus on the issue of preserving the diversity of the generated trajectories. Several heuristic loss functions based on the domain knowledge of basketball are designed to make the generated trajectories assemble real situations in basketball games. Xu et al. in “Contrastive Trajectory Learning for Tour Recommendation” propose a contrastive trajectory learning framework for personalized tour recommendation, which utilizes the intrinsic POI dependencies and traveling intent to discover extra knowledge and augments the sparse data via pre-training auxiliary self-supervised objectives. Chen et al. in “Origin-aware Location Prediction Based on Historical Vehicle Trajectories” propose aTravel TimeDifferenceModel (TTDM), which exploits the difference between the shortest travel time and the actual travel time to predict the next locations by incorporating the travel time from all the passed locations in the query trajectory to each candidate next location. Löffler et al. in “Deep Siamese Metric Learning: A Highly Scalable Approach to Searching Unordered Sets of Trajectories” target fast similarity-based scene retrieval of unstructured ensembles of trajectory data from large databases using Siamese Metric Learning that approximates a distance-preserving low-dimensional representation and that learns to estimate reasonable solutions to the assignment problem.