Classes of Spatio-Temporal Objects and their Closure Properties

  • Authors:

  • Jan Chomicki;Sofie Haesevoets;Bart Kuijpers;Peter Revesz

  • Affiliations:

  • Department of Computer Science and Engineering, University at Buffalo, SUNY, 201 Bell Hall Box 602000, Buffalo, NY 14260-2000, USA e-mail: chomicki@cse.buffalo.edu;Department of Mathematics, Physics and Computer Science, University of Limburg (LUC), Universitaire Campus, B-3590 Diepenbeek, Belgium e-mail: sofie.haesevoets@luc.ac.be;Department of Mathematics, Physics and Computer Science, University of Limburg (LUC), Universitaire Campus, B-3590 Diepenbeek, Belgium e-mail: bart.kuijpers@luc.ac.be;Department of Computer Science and Engineering, University of Nebraska-Lincoln, Ferguson 214, Lincoln, NE 68588-0115, USA e-mail: revesz@cse.unl.edu

  • Venue:
  • Annals of Mathematics and Artificial Intelligence
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present a data model for spatio-temporal databases. In this model spatio-temporal data is represented as a finite union of objects described by means of a spatial reference object, a temporal object and a geometric transformation function that determines the change or movement of the reference object in time.We define a number of practically relevant classes of spatio-temporal objects, and give complete results concerning closure under Boolean set operators for these classes. Since only few classes are closed under all set operators, we suggest an extension of the model, which leads to better closure properties, and therefore increased practical applicability. We also discuss a normal form for this extended data model.