ABSTRACT
Geographic Information System (GIS) is a very useful tool for decision making using spatial data. Spatial decisions are usually made by aggregation of spatial or thematic criteria based on spatial data stored in databases. Spatial data are inherently uncertain, so it is also necessary to aggregate them by appropriately selected functions used for uncertain data. Uncertainty of spatial or thematic criteria can be efficiently expressed by fuzzy sets, and for aggregation of uncertain criteria, fuzzy logic operators (mainly the fuzzy AND, OR, and NOT operators, usually defined as the minimum, maximum, and complement) are commonly used. Other types of aggregation functions are means and averages, which can be also used for aggregation of criteria modelled by fuzzy sets. There is a lot of aggregation operators and each operator has its specific characteristic, so it is important to know which one to use for what purpose. We present the most used aggregation operators such as fuzzy logic operators represented by t-norms and t-conorms, quasi-arithmetic means (or generalized means), a weighted arithmetic mean, and an Ordered Weighted Averaging (OWA) aggregation operator. The result is a description of their implementation in GIS software environments and spatial database systems. The selected operators are applied in solving the tasks of spatial multi-criteria decision making and spatial predictive modelling. The advantages and disadvantages of their particular use are justified in individual cases.
