Spatial Analysis, such as site suitability evaluation and facility site selection, is one of many fields in which the advantages of GIS seem to be broadly accepted in general, if not always in particular. GIS could be satisfied as decision-making ai...
Spatial Analysis, such as site suitability evaluation and facility site selection, is one of many fields in which the advantages of GIS seem to be broadly accepted in general, if not always in particular. GIS could be satisfied as decision-making aid when the problems are well structured. However, facing unstructured problems, for example, the aforementioned task of site evaluation and selection, it is not a satisfactory tool at all.
Many applications of spatial analysis in GIS environment have been accomplished through overlay function, and most commercial GIS overlay methods rely on simple boolean logic. So, adaptation of simple boolean logic to unstructured problem propagate errors and loss of information.
This study examines various aspects of uncertainty with which GIS has to deal, including spatial data and analysis rule uncertainty, spatial variation in land characteristics, and the concept of fuzzy set theory to adapt to Analytic Hierarchy Process for deciding relative weights among factors. The fuzzy set theory and Analytic Hierarchy Process have merits to overcome uncertainty and complexity of spatial data. We can solve much of fuzziness of spatial data and rules with fuzzy set theory, and also, much of complexity of analysis rules with the Analytic Hierarchy Process.
In the evaluations of site suitability, with involves concrete but vague criteria, the Saaty's AHP provides the flexibility of deciding membership functional values calculating membership and selecting functional form with fixed parameters. As for a multi-criteria decision-making, the Saaty's pairs wise comparisons provide a method for classifying, or grouping, the factors. One strategy to deal with complexity is to classify factors into groups. It depends on a common property among factors. Thus one group or class of factors at a certain level of a hierarchy can represent the criteria factors.
The AHP provides a measure of their relative weights on a continuous scale with a capability of grouping and classifying discrete classes and subclasses of factors and consequently a capacity to deal with complex problems. The AHP also generate ratio scale so that analyst can delve into spatial pattern with original data value. The AHP can be economically efficient since we don't have to calculate membership function per each layer, and can be logical since it can represent expert information with consistency.