Calendar Details

For more information about item submission and attendance, see About the Technical Calendar.

Thursday, May 02

Analyzing and Understanding Large and Complex Spatial Interaction Data Using a Graph-theoretic Approach

Caglar, Koylu, University of South Carolina, Columbia,
Geographic Information Sciences and Technology Group Seminar
10:00 AM — 11:00 AM, Research Office Building (ROB), Building 5700, Room E-104
Contact: Budhendra Bhaduri (bhaduribl@ornl.gov), 865.241.9272

Abstract

Moving objects such as people, animals, diseases, flights, commodities, money, and information are among the most influential factors that alter the dynamics of the world such as economy, environment, society, and epidemics. These moving objects represent spatial interactions between places on the earth's surface. Spatial interactions naturally form a location-to-location network (graph) in which a node represents a location (or an area) and a link represents an interaction (flow) between two locations. Spatial interaction patterns exist over multiple spaces (e.g., geographic space, network space, multivariate space) and time. It is a challenging task to discover interesting and unknown patterns within a large dataset of spatial interactions. This research presentation focuses on two new graph-theoretic approaches to discovering spatial interaction patterns that exist in multiple spaces. First, it introduces a new approach to smoothing locational measures to discover the structural characteristics and the roles of locations in spatial interaction networks. When applied to a locational measure such as net migration rate, the smoothing approach can help discover natural regions of attraction (or depletion) and other structural characteristics that the original (unsmoothed) measures fail to reveal. Second, it introduces a measure of social connectedness that summarizes the relationships in a social network embedded in space and time. By mapping the measure values over a series of space-time windows, one can examine the changing dynamics of social relationships across space and time. Such information is crucial for various fields such as disaster evacuation planning and provision of care to elderly.