Travel time estimation (TTE) aims to predict travel duration and provide reliable planning for residential travel schedules. Trajectories naturally contain sequential features in form of GPS points with temporal precedence, which can be leveraged to improve prediction performance. Besides, the spatial information, i.e., the graph structure of the road network, can well represent the road highly and is commonly used to capture spatial information in traffic networks. However, extracting regional spatial information from trajectory data, in addition to its latitude and longitude information, poses a significant challenge due to the inherent format in which the trajectory data is recorded. In light of this, we propose a graph-transformer for TTE (GT-TTE) to utilize a Graph Transformer to adapt effectively to trajectories’ sequential and spatial characteristics for improved TTE performance. By traversing the trajectory nodes with GT-TTE, we construct a graph structure for all trajectory points, thereby obtaining the relative spatial information of each point. Further, we obtain a region adjacency empirically more feature-rich over the sequential data. We evaluate GT-TTE on three real-world representative data sets and observe improvement by approximately 17% compared to the state-of-the-art baselines.

In modern traffic management, one of the most essential yet challenging tasks is accurately and timely predicting traffic. It has been well investigated and examined that deep learning-based Spatio-temporal models have an edge when exploiting Spatio-temporal relationships in traffic data. Typically, data-driven models require vast volumes of data, but gathering data in small cities can be difficult owing to constraints such as equipment deployment and maintenance costs. To resolve this problem, we propose TrafficTL, a cross-city traffic prediction approach that uses big data from other cities to aid data-scarce cities in traffic prediction. Utilizing a periodicity-based transfer paradigm, it identifies data similarity and reduces negative transfer caused by the disparity between two data distributions from distant cities. In addition, the suggested method employs graph reconstruction techniques to rectify defects in data from small data cities. TrafficTL is evaluated by comprehensive case studies on three real-world datasets and outperforms the state-of-the-art baseline by around 8 to 25 percent.