Although GIS and webmapping softwares provide technical means for producing thematic
maps, the selection of relevant visual variables for displaying geographic data is a
task that GIS practionners often do without applying the commonly agreed semiological
rules.
This paper presents the implementation of these graphical rules (Bertin 1967, Hussy
1995, Rod 2000) in a semiological decision support systems based on Mapserver,
PostGIS and PHP technologies. The potential visual variables are sorted by order of
relevance, while the final choice is left to the users.

In thematic maps the choice a visual variable depends on its properties, as well as
on the type of data to be mapped. For instance, a visual variable such as color
allows the selection and the association of objects, but not their ordering (Bertin
1967). Therefore, color should only be used for representing qualitative or nominal
data, such as a soil type, and not for quantitative data like population counts.
Combining the 6 visual variables (shape, orientation, color, texture, value, size)
with the 3 spatial types (point, line, area) produces the generic graphical matrix
containing the 18 basic types of representation (point in color, line in color, area
in color, point in texture, etc.).
The proposed semiological decision support system provides the user with an
evaluation of the most suitable variable visual to be used for mapping the available
data. The system is based on metadata describing the data to be mapped (data types,
spatial units) and tries to match the nature of the selected data with the properties
of the potential visual variables. The final decision for a specific visual variable
is left to the user.
The system is implemented using the follwing techonolgies: PostGIS (metadata and
geodata), Mapserver (webmapping), PHP/javascript and HTML (scripts, user’s interface).

Bibliographic references

Andrienko G., Andrienko N., « Interactive maps for visual data exploration »,
International Journal Geographical Information Science, vol. 13, 1999, p. 355-374.

Bertin J., Sémiologie graphique, Paris, Mouton, 1967.

Brewer C., ColorBrewer, 2002, http ://www.personal.psu.edu/faculty/c/a/cab38/
ColorBrewerBeta.html.

Cauvin C., Reymond H., Serradj A., Discrétisation et représentation cartographique,
Montpellier, RECLUS, Modes d’emploi, 1987.

Dykes J. A., « Cartographic Visualization : Exploratory Spatial Data Analysis with
Local Indicators of Spatial Association using Tcl/Tk and cdv », The Statistician,
47(3) , 1998, p. 485-497.

Hussy C., La carte, un modèle, un langage, Genève, Dépt de Géographie, 1995.
Jung V., « Knowledge-based Visualization Design for Geographic Information Systems »,
Proc. 3rd ACM Workshop on Advances in GIS, ACM Press, 1995, p.101-108.

MacEachren A.M., How maps work – Representation, Visualization and Design, New York,
The Guilford Press, 1995.

MacEachren A.M., Kraak M.-J.,  « Exploratory cartographic visualization : advancing
the agenda », Computers and Geosciences, vol. 23, 1997, p. 335-344.

Mackinlay J., « Automating the Design of Graphical Presentation of Relational
Information », ACM Transactions on Graphics 5, 1986, p. 110-141.

Rod J. K., « The third choice », Cybergeo, No 154, 2000, http
://193.55.107.45/semiogra/ rod/rod.htm.

Weibel R., Buttenfield B., « Improvement of GIS graphics for analysis and
decision-making », International Journal of Geographical Information Systems., vol.
6(3), 1992, p. 223-245.

Zhan F. B., Buttenfield B., « Object-Oriented Knowledge-Based Symbol Selection for
Visualizing Statistical Information », International Journal of Geographical
Information Systems, vol. 9(3) , 1995, p. 293-315.