• DocumentCode
    1365574
  • Title

    Design Study of LineSets, a Novel Set Visualization Technique

  • Author

    Alper, Basak ; Riche, Nathalie Henry ; Ramos, Gonzalo ; Czerwinski, Mary

  • Volume
    17
  • Issue
    12
  • fYear
    2011
  • Firstpage
    2259
  • Lastpage
    2267
  • Abstract
    Computing and visualizing sets of elements and their relationships is one of the most common tasks one performs when analyzing and organizing large amounts of data. Common representations of sets such as convex or concave geometries can become cluttered and difficult to parse when these sets overlap in multiple or complex ways, e.g., when multiple elements belong to multiple sets. In this paper, we present a design study of a novel set visual representation, LineSets, consisting of a curve connecting all of the set´s elements. Our approach to design the visualization differs from traditional methodology used by the InfoVis community. We first explored the potential of the visualization concept by running a controlled experiment comparing our design sketches to results from the state-of-the-art technique. Our results demonstrated that LineSets are advantageous for certain tasks when compared to concave shapes. We discuss an implementation of LineSets based on simple heuristics and present a study demonstrating that our generated curves do as well as human-drawn ones. Finally, we present two applications of our technique in the context of search tasks on a map and community analysis tasks in social networks.
  • Keywords
    cartography; data visualisation; geometry; set theory; community analysis task; concave geometry; convex geometry; heuristics; information visualization; linesets; map; search tasks; set visual representation; set visualization; social network; Accuracy; Data visualization; Geometry; Shape analysis; Social network services; Set visualization; clustering; faceted data visualization; graph visualization.;
  • fLanguage
    English
  • Journal_Title
    Visualization and Computer Graphics, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    1077-2626
  • Type

    jour

  • DOI
    10.1109/TVCG.2011.186
  • Filename
    6064991