Archive for the ‘Uncategorized’ Category

Information Visualization and Proxemics: Design Opportunities and Empirical Findings

Information visualization uses interactive graphics to amplify cognition. It can improve many aspects of dealing with large sets data: Visualizations help explore and navigate large information spaces, analyze and make discoveries in high-dimensional data, and discuss data within on-line communities.

Most information visualizations—commercial products and research prototypes alike—are designed for a setting where the user interacts using a mouse on a desktop-sized display. Recent research has explored how visualizations should be designed for non-desktop settings, in particular for large high-resolution displays. Examples of visualizations designed for this setting include using tangible input controllers, sensing body movements as implicit navigation input, and adapting interaction techniques for large displays.

We extend this work by using the notion of proxemics to identify design opportunities. Proxemics studies the relation between people as it is expressed in the use of space. Compared to early work on proxemics, recent work as well as this paper extend the notion of proxemics to describe also the relation between people and objects (often user interfaces). In research on human-computer interaction (HCI), proxemics has for instance been used to design interaction techniques that change user interface layout based on the user’s position, and to study orientation and distance among devices and doctors in neurosurgery. Previous research has also demonstrated how body orientation and position can be used with visualizations: for implicit interaction with ambient displays and for coarse 3D navigation in microseismic visualizations. We build on previous work to explore how the notion of proxemics can be applied to interaction with information visualization.

The opportunities for proxemics in information visualization are manifold. First, it may be used to adapt visualizations based on the users’ position and orientation relative to the display. Second, it could use movements in front of a display to have visualizations follow users’ movements or blend as two users get close. Third, we could augment users’ backing away from a large display by even further zooming out or abstracting the visualizations. Many other uses of proxemics in information visualizations may be imagined.

This paper explores in particular design opportunities for information visualization based on movement and distance to large high-resolution displays. We focus on using movement and distance because earlier work has emphasized physical navigation as important when using large displays and in group work. We explore spatial relations only between a single user and visualizations; exploring relations between people would provide more opportunities, but is beyond the scope of this paper. The opportunities are illustrated with a design space and with sketches; the opportunities focus both on supplementing other input techniques and on replacing them. We also show how earlier work that has not explicitly used the notion of proxemics can be understood through proxemics and potentially benefit from its analytic framework. We select a subset of design opportunities to implement and test in three user studies:

(1) navigation by physical movement,

fig5.1afig5.1bfig5.1cfig5.1dfig5.1efig5.1f

(2) querying coordinated views by movement, and

fig5.3a fig5.3b fig5.3c fig5.3d

(3) adapting visual representations to distance.

fig5.2a fig5.2b fig5.2c

We do so to generate design ideas, but also to provide initial data on the usefulness of combining information visualization and proxemics. Our approach is to ground some opportunities in empirical data rather than to give an exhaustive systematic review of the opportunities or to present indepth data on a single case.

We contribute (a) an initial analysis of using proxemics for information visualization, (b) prototypes of information visualizations that adapt based on tracking of their users, and (c) an evaluation of a set of proxemic visualizations. The argument is that proxemics may offer promising design opportunities for non-desktop visualizations; we think such opportunities are valuable to both researchers in visualization and to designers for large displays.

Publication

paperJakobsen, M. R., Haile, Y. S., Knudsen, S., Hornbæk, K. 2013. Information Visualization and Proxemics: Design Opportunities and Empirical Findings. In IEEE Transactions on Visualization and Computer Graphics. vol. 19, no. 12, Dec. 2013.

An Exploratory Study of How Abundant Display Space May Support Data Analysis

Last week I presented a study at NordiCHI’12 which I conducted in collaboration with my colleagues Mikkel R. Jakobsen and Kasper Hornbæk last winter.

The study is based on 11 workshops in which we asked participants, which were groups of 2 to 3 analysts from diverse domains of data analysis, to imagine, and act as if the whiteboard in front of them was a large interaction high-resolution display. The groups imagined using the display to work with their own tasks and data to assure that their imagination was bound to concrete and meaningful work.

It seemed that the presentation was well received in the audience. The questions from the audience related to

  • Private areas, public areas and parking areas as previously described in the litterature, primarily in relation to tabletop surfaces
  • The limitation of the interaction technologies (i.e. using magnets to mimick a dragging interface etc.)

Feel free to leave comments related to the presentation or the paper.

S. Knudsen, M. R.  Jakobsen & K. Hornbæk. An Exploratory Study of How Abundant Display Space May Support Data Analysis, in Proceedings of NordiCHI’12, 2012