A spatiotemporal sequence mining approach for ship trajectory data

Authors: Jing Li*, University of Denver, Tong Zhang, Wuhan University, Xuantong Wang, University of Denver
Topics: Geographic Information Science and Systems, Cyberinfrastructure
Keywords: Spatiotemporal data mining, ship movements, visualization, Automatic Identification System (AIS)
Session Type: Paper
Day: 4/6/2019
Start / End Time: 8:00 AM / 9:40 AM
Room: Madison A, Marriott, Mezzanine Level
Presentation File: No File Uploaded


With the development of sensor techniques, fine level tracking data become increasingly accessible. One example is the tracking information of ships gathered from the Automatic Identification System (AIS). Real time and historical AIS data have been intensively used to various scenarios such as abnormal patterns detection, route predictions and collision detection. Very few studies have conducted the efficiency analysis of maritime transportation operations on such data. This paper describes a sequence pattern based mining method pertaining to port operations, movements of ships and design of waterways. The core method is the sequence pattern detection based on trajectory data. The method consists of four steps: 1) reconstruction of trajectories using tracking data from AIS; 2) determination of spatiotemporal conditioned region of interests (ROIs) and the association with port locations; 3) generation of visual representations of regional maritime transportation status using ROIs and simplified representations of trajectories; In all steps, we incorporate interactive tools to facilitate the customization and the configuration of the steps to engage users in the analysis process. Typical visualization techniques such as interactive filtering, summarization, sensitive analysis are used. This method can facilitate the evaluation of efficient maritime transportation operations. To demonstrate the feasibility of the method, we have created a web based prototype system and conducted tests using AIS data for US coastal regions.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login