Seeking structure in records of spatio-temporal behaviour: visualization issues, efforts and applications

  • Authors:

  • J. A. Dykes;D. M. Mountain

  • Affiliations:

  • Department of Information Science, City University, Room A217a Northampton Square, London EC1V 0HB, UK;Department of Information Science, City University, Room A217a Northampton Square, London EC1V 0HB, UK

  • Venue:
  • Computational Statistics & Data Analysis - Data visualization
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

Information that contains a geographic component is becoming increasingly prevalent and can be used both to analyse relatively complex behaviours in time and space and to combat the potential for information overload by assessing the geographic relevance of information. Such analysis can be combined with mobile communications technology to fuel location-based services that offer information pertinent in terms of geography, time, experience and preference. This paper aims to raise some issues relating to these advances and describes novel representations designed for interactive graphical exploratory data analysis (EDA). A number of graphical techniques and representation methods are introduced to establish the nature of the kinds of data that are being collected and the suitability of visualization for EDA of spatio-temporal data. These include the interactive views provided by the Location Trends Extractor, 'spotlights'--continuous density surfaces of recorded spatio-temporal activity, networks of morphometric features derived from continuous surfaces representing density of activity and geocentric parallel plots presented in a spatial multimedia environment for data exploration. Some of the benefits and limitations of the techniques are outlined along with suggestions as to how the visualization tools might be utilized and developed to improve our understanding of behaviour in time and space and evaluate and model geographic relevance.