Tutorial 3
Similarity-based analytics for trajectory data: theory, algorithms and applications
Kai Zheng
The prevalence of GPS sensors and mobile devices has enabled tracking the movements of almost any kind of moving objects such as vehicles, humans and animals. As a result, in the past decade we have witnessed unprecedented increase of trajectory data both in volume and variety. With some attributes such as variable lengths, uncontrolled quality, high redundancy and uncertainty and so on, trajectory data challenge the traditional methodologies and practices in many research areas including data storage and indexing, data mining and analytics, information retrieve, etc. Trajectory data management has been attracting numerous research interests from both academia and industry due to its tremendous value and benefits in a variety of critical applications like traffic analysis, fleet management, trip planning, location-based recommendation, etc. In this tutorial, we will talk about the challenges, techniques and open problems with the focus on similarity-based analytics, the foundation of trajectory management, and covering a range of topics from fundamental theory, algorithms to advanced applications.