Evaluating event visualization: a usability study of COPLINK spatio-temporal visualizer

  • Authors:

  • Wingyan Chung;Hsinchun Chen;Luis G. Chaboya;Christopher D. O'Toole;Homa Atabakhsh

  • Affiliations:

  • Department of Information and Decision Sciences, Colleges of Business Administration, The University of Texas at El Paso, 500 W, University Avenue, El Paso, TX;Artificial Intelligence Lab, Department of Management Information Systems, The University of Arizona, 1130 East Helen Street, McClelland Hall 430, Tucson, AZ;Artificial Intelligence Lab, Department of Management Information Systems, The University of Arizona, 1130 East Helen Street, McClelland Hall 430, Tucson, AZ;Artificial Intelligence Lab, Department of Management Information Systems, The University of Arizona, 1130 East Helen Street, McClelland Hall 430, Tucson, AZ;Artificial Intelligence Lab, Department of Management Information Systems, The University of Arizona, 1130 East Helen Street, McClelland Hall 430, Tucson, AZ

  • Venue:
  • International Journal of Human-Computer Studies
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Event visualization holds the promise of alleviating information overload in human analysis and numerous tools and techniques have been developed and evaluated. However, previous work does not specifically address either the coordination of event dimensions with the types of tasks involved or the way that visualizing different event dimensions can benefit human analysis. In this paper, we propose a taxonomy of event visualization and present a methodology for evaluating a coordinated event visualization tool called COPLINK Spatio-Temporal Visualizer (STV). The taxonomy encompasses various event dimensions, application domains, visualization metaphors, evaluation methods and performance measures. The evaluation methodology examines different event dimensions and different task types, thus juxtaposing two important elements of evaluating a tool. To achieve both internal and external validity, a laboratory experiment with students and a field study with crime analysis experts were conducted. Findings of our usability study show that STV could support crime analysis involving multiple, coordinated event dimensions as effectively as it could analyze individual, uncoordinated event dimensions. STV was significantly more effective and efficient than Microsoft Excel in performing coordinated tasks and was significantly more efficient in doing uncoordinated tasks related to temporal, spatial and aggregated information. Also, STV had compared favorably with Excel in completing uncoordinated tasks related to temporal and spatial information, respectively. Subjects' comments showed STV to be intuitive, useful and preferable to existing crime analysis methods.