POWRE: Extend Geographic Representation to Support Knowledge Discovery in Geographic Information Systems

University Of Oklahoma Norman Campus, Norman OK

Investigators

Abstract

This POWRE project will provide the resources needed to study representations of dynamic geographic processes and methods of geospatial knowledge discovery about patterns and behaviors of these processes in Geographic Information Systems (GIS). The study will address three key research questions: (1) What kinds of information are useful to understand geographic processes? (2) How should geographic processes be represented so that needed information can be computed from GIS databases? and (3) How can exploratory and analytical tools be designed with the proposed geographic representation to facilitate knowledge discovery in GIS? Research on representation for dynamic geographic processes is urgently needed and can have a significant impact on methodologies for all sciences that use geospatial data to understand the dynamics of physical and human systems. As GIS development gradually merges into mainstream information technology (IT), innovative GIS research that emphasizes the computational fundamentals of geographic information representation and analysis requires advanced knowledge and skills in a melange of GIS, IT, and a chosen application domain. To meet such a multidisciplinary challenge, the POWRE study will start a new line of research that applies data mining and knowledge discovery methods in GIS to unveil spatial patterns of weather behaviors from meteorological and climatological data. The research framework to be developed in the study will be readily applicable to GIS knowledge discovery in other application domains. The POWRE award will give the PI, recently granted academic tenure, a unique opportunity to develop a major research program in geographic representation and knowledge discovery in GIS at this juncture in her career. This ultimately will contribute to a greater integration of GIS and IT technologies to enhance the usefulness of volumetric scientific data that are costly acquired at a phenomenal rate.