With the development of navigation globalization and ship dehumanization, the contradiction between the increasing demand for ship behavior supervision and limited traffic service resources is obvious, and the frequent occurrence of accidents at sea is a problem. The monitoring of abnormal ship behavior is an important link in maritime transportation. With the popularization of the automatic identification system and increasing research in the maritime field, the AIS is widely used in the management of ship static information and the real-time sharing of dynamic information. The generated moving ship trajectory data provide a new opportunity for research into abnormal ship behavior and its detection. In light of the current situation of abnormal ship behavior research, we detected abnormal ship behavior from the point of view of spatial information and thematic information based on moving ship trajectory data. Therefore, this study first modeled the cognition of abnormal ship behavior. Then, based on the cognition of group ship behavior rules, we used a method based on graph structure learning to mine maritime routes from the perspective of ship spatial information. Next, we used Rayda’s criterion to detect the anomalous behavior of ships in space. Then, based on the isolation forest algorithm, we detected and described the abnormal behavior shown by ship thematic information. The experimental results show that the framework proposed in this paper can effectively detect the abnormal behavior of ships.

Traffic congestion detection poses challenges in spatiotemporal data mining and intelligent transportation research. Existing studies primarily detect traffic congestion based on the speed estimation of traffic flows. Such detection techniques may overlook the formation of traffic congestion in space and time. This research proposes a density-based approach to moving object clustering that extracts the spatiotemporal extents of traffic congestion in three steps. The first step applies a map-matching strategy to project original trajectory points in a planar space onto a road network space and segments the trajectories into consecutive time windows. In the second step, we statistically detect moving clusters with significantly high-density subject to network constrained clustering. The final third step determines moving clusters indicative of traffic congestion through the analysis of both vehicle speed and time spans. Comparative experiments on both simulated trajectories and the real-life taxi trajectories in Wuchang demonstrate that the proposed method outperforms other methods through quantitative evaluations using three indicators, i.e. the precision, recall and F1 value. The proposed approach can illustrate the spatiotemporal regularities of traffic congestion, which can inform dynamic route planning and network design optimization.

Existing spatial clustering methods primarily focus on points distributed in planar space. However, occurrence locations and background processes of most human mobility events within cities are constrained by the road network space. Here we describe a density-based clustering approach for objectively detecting clusters in network-constrained point events. First, the network-constrained Delaunay triangulation is constructed to facilitate the measurement of network distances between points. Then, a combination of network kernel density estimation and potential entropy is executed to determine the optimal neighbourhood size. Furthermore, all network-constrained events are tested under a null hypothesis to statistically identify core points with significantly high densities. Finally, spatial clusters can be formed by expanding from the identified core points. Experimental comparisons performed on the origin and destination points of taxis in Beijing demonstrate that the proposed method can ascertain network-constrained clusters precisely and significantly. The resulting time-dependent patterns of clusters will be informative for taxi route selections in the future.

The estimated time of arrival (ETA) plays a vital role in intelligent transportation systems and has been widely used as a basic service in ride-hailing platforms. Obtaining a precise ETA is a challenging task due to the complexity of the real-world geographic and traffic environments. Previous works suffer from heterogeneous sparse data learning and multiple-correlation extraction issues. Therefore, this paper presents a hybrid deep learning framework (HSETA) to estimate the vehicle travel time from massive data. First, we encode heterogeneous data to represent various features in different respects. Then, we develop an ensemble factorization machine block (EFMB) structure combined with gated recurrent unit (GRU) and multilayer perceptron (MLP) to extract information from sparse and dense features. Next, the multiple-correlation learning block (MCLB) structure that we propose is utilized to aggregate information based on multiple correlations. Finally, the travel time can be estimated by simple regression. Our extensive evaluations on two real-world datasets show that HSETA significantly outperforms all baselines. Our PyTorch implementation of HSETA and sample data are available at https://github.com/LouisChenki/HSETA