Data Management and Analysis on Taxi Trajectories

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "Data Management and Analysis on Taxi Trajectories"

By

Mr. Xibo ZHOU


Abstract

With the ubiquity of location sensing technologies in a wide range of 
location-based services such as GPS navigators, large amounts of 
trajectory data have been collected. These data have been utilized by 
various applications such as location-based services, urban planning, and 
human behavior analysis. As a major type of location-based services, taxis 
have become an important part of the public transportation system in many 
large cities, providing convenience for our daily life. In practice, the 
information contained in taxi trajectory data are imprecise and incomplete 
due to various factors such as measurement noises, low sampling rate, and 
sparsity. In this thesis, we study the problem of data calibration and 
applications of knowledge discovery from the perspective of three 
important features of taxi trajectory data, namely location information, 
occupancy status, and travel speed. From each perspective, we study and 
propose a specific application regarding to the feature, including: 
1)interactive map-matching for location calibration of trajectories; 
2)taxi fraud detection on top of occupancy status prediction; and 3) 
speeding prediction on low-sampled and sparse taxi trajectories.

For the interactive map-matching problem, we design a framework that 
combines efforts with algorithms in an interactive manner to achieve high 
map-matching accuracy, and propose various query selection strategies to 
effectively reduce the annotation cost. For taxi fraud detection problem, 
we introduce the a new type of taxi fraud called unmetered taxi rides, and 
propose a learning model to predict the passenger occupancy status of 
taxis, and implement a heuristic algorithm to find fraudulent 
trajectories. For the taxi speeding prediction problem, we propose a 
learning model to predict the travel speed of individual taxis, which is 
applied for detecting taxi speeding.


Date:			Friday, 19 January 2018

Time:			2:00pm - 4:00pm

Venue:			Room 3494
 			Lifts 25/26

Chairman:		Prof. Chik Patrick Yue (ECE)

Committee Members:	Prof. Lionel Ni (Supervisor)
 			Prof. Qiong Luo (Supervisor)
 			Prof. Lei Chen
 			Prof. Shing-Chi Cheung
 			Prof. Jingshen Wu (MAE)
 			Prof. Qing Li (Comp. Sci., CityU)


**** ALL are Welcome ****