Developing a Novel Framework for Analyzing Dynamic Interactions in Movement Data

点击翻译按钮获取中文摘要

Technological advancements in global positioning systems and related satellite tracking technologies have resulted in significant increases in the availability of highly accurate data about moving objects. These advances, however, have dramatically outpaced the development of appropriate methods for analyzing movement data. This research project will develop a benchmarking framework for measuring and interpreting dynamic interactions based on both temporal and spatial criteria among individual entities such as people or animals. The project will advance the spatial study of movement beyond the identification of the location of individual entities as specific objects and consider interaction as a variable for which a basic unit of observation is a set of locations with time stamps for multiple individual entities. Analyses of interactions, for example, can provide more information about possible attraction and avoidance of individuals that are in the same area at the same time and can be useful for understanding how two individuals interact in the context of disease transmission and social behavior. While this research focuses on animal movement and interactions, the results will be applicable to other moving objects that interact. Project results therefore could potentially be used to study a wide variety of phenomena, including disease spread, information diffusion, crowd movement, and traffic behavior. The project also will promote education and training through support of student research and development of curricular materials. This research will involve the conduct of an extensive comparison of existing dynamic interaction metrics. The investigator will examine their strengths and weaknesses, how they are affected by temporal resolution, spatial uncertainty, and level of interaction. The ability of these metrics to quantify interactions will be assessed using two simulated movement datasets as well as one high-quality real dataset for which the level of interaction is known. This comparison will contribute recommendations and guidelines for research that aims to quantify interactions among animals, people, automobiles, or other moving objects. The results from the comparison also will be used to develop new interaction metrics that will be more appropriate for measuring context- or behavior-specific types of interactions.