Spatial Decision Support Systems

Origins of SDSS

The interest in SDSS dates back to the late 1980s, when GIS was becoming an important information technology for automating, managing, and analyzing geographic data. Analytical tools of GIS incorporating map overlay, network analysis, attribute data manipulation, and attribute and spatial query were sufficient to find locations satisfying multiple decision criteria. However, GIS lacked tools for predicting impacts of choosing a specific location site on criteria that might have been of interest to decision makers but were not present in GIS database. Such problems are called “ill-structured” or “partially structured” because the bases for making a decision are not well-known or cannot be foreseen ahead of a decision problem. To overcome the difficulty of dealing with partially structured problems, the initial idea of SDSS included a model database, from which problem-specific impact models could be accessed to compute impacts of various decision options. Later, in the course of the 1990s, the idea of model base was replaced by a simpler and more pragmatic approach, in which specific impact models were built into SDSS once the decision criteria serving as the bases for evaluating decision options became known. More recently, the idea of a model database storing model components has come to the fore again with the development of interoperable Web services that allow impact models provided as independent Web services to access various spatial databases distributed on the Internet.

SDSS and Decision Process

The most fundamental activity of decision making involves a selection (choice). Selection is often preceded by ordering decision options from best to worst or by sorting them into categories such as good, average, or unacceptable. Choosing a course of action is synonymous with making a decision. Hence, decision-making activity requires at least two decision alternatives: two feasible solutions to a decision problem. Decision making is not an act, as it erroneously sometimes seems to be perceived, but a process, which aims at resolving three questions:

• What is the problem that calls for making a decision?
• What are the decision alternatives (options)?
• Which alternative is the best?

[Page 407]Solutions to spatial decision problems are characterized by multiple criteria upon which they are judged (evaluated). Consider an example of a common spatial decision problem tackled with the help of GIS: a site selection. Finding suitable locations involves the search of a potentially large number of candidate locations using suitability criteria, for example, distance to roads, land cost per acre, zoning designation, soil drainage characteristic, and so on. Suitability criteria are here the basis for site selection.

...