Theme 5: Representations

This theme introduces students to a wealth of (visual) representations. As there are a variety of well-known techniques for representing data, and because it is impossible to present an exhaustive list, effective categorizations of presentations are important.

(5.1) General discussion

Many visualization problems can be addressed by choosing from among traditional representations. In other situations, new representations must be designed. In either case, effectiveness and expressiveness of the resulting image are crucial. Expressiveness refers to the ability of a visual display to represent (i.e., to encode) the data. Effectiveness refers to the ease with which a user can interpret (i.e., to decode) the representation. Use the dictum first, make sure the picture is accurate.

(5.2) Computer Graphics

Many issues of Computer Graphics are pertinent to understand (power and pitfalls of) visualization techniques. For the design of new visualization techniques often an in-depth comprehension of computer graphics algorithms is necessary.

(5.3) Selection Criteria

To make the right choice among the offered visualization techniques, selection criteria must be known and used correctly. These include characteristics of the data (such as type, dimensionality, structure or topology), the purpose of the visualization (which might be exploration, confirmation, or presentation), the visualization goal and output medium.

(5.3.1) Several factors must be considered when choosing representations, including

(5.3.1.1) Characteristics of the data, such as

(5.3.1.1.1) Type

(5.3.1.1.1.1) Is the data nominal, ordinal, or quantitative? For quantitative data, is it a scalar, vector, or tensor quantity?);

(5.3.1.1.2) Spatial characteristics

(5.3.1.1.2.1) Is the data geographical?

(5.3.1.1.3) Dimensionality of the data

(5.3.1.1.4) Temporal characteristics

(5.3.1.1.5) Topology of the data

e.g. is the data scattered, unstructured, structured?

(5.3.1.2) The purpose of the visualization

which might be exploration, confirmation, or presentation. Presentation could be to a small group of colleagues or to a larger audience.

(5.3.1.3) The [user / goal / interpretation] aim of the visualization.

(5.3.1.4) The output medium [paper, video, interactive?]

(5.4) Techniques

This chapter describes a number of well-known techniques for representation in order to suggest the variety of representations that have been used in visual data representation without attempting to offer an exhaustive list. Describe first single techniques, such as line graphs, scatterplots, glyphs, contour plots, streamlines, isosurfaces, or cone trees, then several organizational structures of techniques, such as animation, fish eye view, or linked maps.

(5.4.1) Describe number of well-known techniques for representation

This chapter suggests the variety of representations that have been used in visual data representation rather than attempting to offer an exhaustive list:

(5.4.1.1) line graph, histogram, bar chart, strip chart, pie chart

(5.4.1.2) scatterplots

(5.4.1.3) glyph and icons, e.g. Chernoff faces, Andrews Plots, icons

(5.4.1.4) parallel coordinates

(5.4.1.5) map, contour plot, surfaces (height fields), raster images

(5.4.1.6) particles, streamline, streakline

(5.4.1.7) isosurface, direct-volume rendering

(5.4.1.8) cone trees

(5.4.1.9) ball-and-stick model

(5.4.2) Organizational Structure of representations

Representations are presented within an overall context. For example, two isosurfaces might be presented side-by-side to facilitate comparison between two data sets. Or, a series of raster images might be presented in a row, each one showing a different frequency band from satellite imagery of some area. Other structures include:

Menü