Geographic information is an indispensable resource for spatial decision making. With the wide availability of new information sources, such as volunteered geographic information and sensor networks, decision makers nowadays are often overwhelmed by a large volume of information. This information deluge presents a challenge for not only professional decision makers but also general individuals who need to make decisions for a variety of daily tasks. Such a challenge calls for a method that can quantify the value of geographic information, so that decision makers can be firstly informed about the more important information. Information value theory (IVT) from economics and artificial intelligence has provided some guidance on this issue. However, existing IVT research has mostly focused on monetary values, while the spatiotemporal factors contained in geographic information have been largely neglected.

The main contributions of this dissertation are extensions of information value theory in the context of human mobility. Specifically, this research aims at understanding the role of geographic information in facilitating the movement planning of individuals or small groups, and therefore presents a micro-scale perspective on human mobility. This dissertation is developed based on three peer-reviewed articles of the author, which have been arranged as chapter 3, 4, and 5 respectively. Chapter 3 presents a geo-ontology design pattern for semantic trajectories, which serves as a foundational data model for linking, reasoning, and analyzing movement data. Chapter 4 and 5 extend IVT in two different applications that involve human movements. Chapter 4 values geographic information with regard to the spatiotemporal tasks of individuals in daily life, and integrates IVT with space-time prisms to quantify the value of information.

Chapter 5 investigates the value of geographic information in disaster responses, and quantifies information value based on its capability of assisting emergency responders in planning relief trips. Methodological details are presented, and a variety of simulations and real-world cases are employed to examine the performances of the proposed methods and frameworks. While this work has its limitations, I hope it can make a modest contribution towards understanding and quantifying the value of geographic information.