• Jun 17] Visual Storytelling Contest videos and materials are updated.
• Apr 14] Workshops (PacificVAST) schedules are updated.
• Mar 7] Conference program and keynote talks are updated.
• Feb] Registration is now available!
• Feb] Accommodation information is updated.

• 2016
  Taipei
  Apr 19 - Apr 22
• 2015
  Hangzhou
  Apr 14 - Apr 17
• 2014
  Yokohama
  Mar 04 - Mar 07
• 2013
  Sydney
  Feb 26 - Mar 01
• 2012
  Songdo
  Feb 28 - Mar 02
• 2011
  Hong Kong
  Mar 01 - Mar 04
• 2010
  Taipei
  Mar 02 - Mar 05
• 2009
  Beijing
  Apr 20 - Apr 23
• 2008
  Kyoto
  Mar 04 - Mar 07

• Facebook Official Facebook Page

• Twitter Official Twitter Account

• WeChat Official WeChat Account (亚太地区可视化会议微信公众号)

Towards Perceptual Optimization of the Visual Design of Scatterplots

• Luana Micallef luana.micallef (at) hiit.fi
  Helsinki Institute for Information Technology HIIT, Aalto University, Helsinki, Finland
• Gregorio Palmas gpalmas (at) kth.se
  KTH Royal Institute Of Technology, Stockholm, Sweden
• Antti Oulasvirta antti.oulasvirta (at) aalto.fi
  Aalto University, Helsinki, Finland
• Tino Weinkauf weinkauf (at) kth.se
  KTH Royal Institute Of Technology, Stockholm, Sweden

Abstract

Designing a good scatterplot can be difficult for non-visualization experts, because they need to decide many parameters, such as marker size and opacity, aspect ratio, color, and rendering order. This paper contributes to research exploring the use of perceptual models and quality metrics to automatically set such parameters to enhance the visual quality of a scatterplot. A key consideration in this paper is the construction of a cost function to capture several relevant aspects of the human visual system examining a scatterplot design in some data analysis task. We show how the cost function can be used in an optimizer to search for the optimal visual design for a user窶冱 dataset and task objectives (e.g., 窶腕eliable linear correlation estimation is more important than class separation窶). The approach is extensible to different analysis tasks. To test its performance in a realistic setting, we pre-calibrated it for correlation estimation, class separation, and outlier detection. The optimizer was able to produce designs that achieved a comparable level of speed and success with those using human-designed presets (e.g., R, Matlab). Case studies demonstrate that the approach can adapt a design to the data to reveal patterns without user intervention.