Spatio-Temporal Data Mining and Analyses in a Multi-Scale Framework

Authors: Yi Qiang*, University of Hawaii - Manoa
Topics: Spatial Analysis & Modeling, Geographic Information Science and Systems, Land Use and Land Cover Change
Keywords: spatio-temporal, data mining, multi-scale, pattern detection
Session Type: Paper
Day: 4/5/2019
Start / End Time: 1:10 PM / 2:50 PM
Room: Roosevelt 3, Marriott, Exhibition Level
Presentation File: No File Uploaded

The representations of space and time are fundamental to GIScience and data mining of diverse geospatial data collected at different resolutions. The choice of analytic scales (both spatial and temporal) to a large extent determines the insights that can be gained. The importance of scale has been epitomized in the well-known modifiable areal unit problem (MAUP) and its temporal equivalent. However, multi-scale analysis for spatio-temporal data remain a challenge for prevalent GISystems, which represent geospatial data represented as flat layers at pre-determined resolutions and spatial analysis tools only operate at a single scale. To discover patterns and relationships that are prominent in different scales, inefficient “trial-and-error” approaches are typically used to determine the appropriate resolution, aggregation units and search bandwidth for data analyses and modeling. This presentation will introduce an analytical framework that organize space, time, spatial scale and temporal scale unified and hierarchical structure. By visualizing and analyzing variations of data in the scale dimensions, nested relationships and patterns at multiple spatial and temporal scales can be better quantified and understand. This will also demonstrate a series of applications of the framework in time series analysis, spatial pattern detection, and land cover change modeling.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login