Attribute Explorer

Updated: 02/11/2008, G. Waloszek

Attribute Explorer

Figure 1: Screenshot from a Java standalone prototype in Java by Andy Smith et al. (IBM, UK) (by the author); note the highlighted house that is being brushed to linked histograms

 

Purpose

Exploration of relations between attributes of multivariate data to gain insight. The technique is based on linked histograms and includes brushing for immediate feedback.

Andy Smith summarized its benefits to include (adapted):

  • An attribute-based display showing the distribution of objects by attribute, and allowing the application of a clearly indicated constraint without eliminating from consideration objects that fail the constraint (see the sliders below the histograms).
  • Multiple concurrent attribute displays (typically histograms), each display being able to show the cumulative effect of all applied attribute constraints.
  • The ability to discern attribute relationships through very rapid feedback in response to the modification of attribute constraints (brushing).

Brushing

Brushing

Figure 2 : Brushing two houses (marked in red) from one plot to the other two... (after Spence, 2007)

According to Spence, brushing is "a change in the encoding of one or more items essentially immediately following, and in response to, an interaction with another item." For illustration, think of objects that can be classified along several dimensions, such as houses that can be classified according to price, number of rooms, and time to drive to work. The house data can be displayed in a 3D-scatterplot, a 2D projection of it, or three 2D projection planes, which show the relations for two of the three dimensions. Users may pick items in one scatterplot and highlight them. Brushing means that the same items in the other two plots are also highlighted immediately. Spence would call this: "We are brushing houses from one plane to the other two." All in all, brushing allows users to explore the effect of changes in one parameter on the relation between the other (here, two) parameters (see figure 2).

Applications

Display and exploration of multivariate data, such as house or car data (examples: auto kiosk application, EZChooser).

See also the EZChooser and the Influence Explorer.

Authors, Date

Robert Spence , Lisa Tweedie (Imperial College of Science, Technology and Medicine, London); 1994-98

Links, Papers

  • Tweedie, L., Spence, R., Williams, D.M.L., & Bhogal, R. (1994). The Attribute Explorer. ACM, Conference Companion Proceedings CHI '94, pp. 436-436. (PDF in ACM Portal)
  • Spence R. & Tweedie, L. (1998). The Attribute Explorer: information synthesis via exploration. Interacting with Computers, 11, pp. 137-146.
  • Robert Spence (2007). Information Visualization (2nd Edition). Prentice-Hall (Pearson).
  • Robert Spence (2002). Sensitivity encoding to support information space navigation:a design guideline. Information Visualization, 1(2), pp. 120-120 (HTML)
  • Robert Spence's Website (last update 2002)
  • Andy Smith (2001): Attribute Explorer - A dynamic query mechanism. (IBM developerWorks; provides a downloadable Java applet; I was not able to open the applet)
  • Andy Smith, Simon Moore, & Ryan Bennitt (2003): Visual Attribute Explorer. (IBM alphaWorks; provides a downloadable stand-alone Java prototype; works in Windows, I was not able to run the version on Mac OS X)
  • Wittenburg, K., Lanning, T., Heinrichs, M., & Stanton, M. (2001). Parallel bargrams for consumer-based information exploration and choice. ACM, Proceedings of UIST '01, pp. 51-60.