Framework for Discovering Association Rules in a Fuzzy Data Cube

摘要

This work proposes a framework to model the geography according how it is presented in the real world, taking advantage of the new information technologies. The concepts in which the multidimensional model is based on and the Fuzzy Logic are complemented for the generation of the Fuzzy Data Cube. Data Mining techniques are applied to the information stored in the FDC, in order to extract the set of association rules, which define the behavior of spatial objects in their environment. Fuzzy sets defined in the FDC, allow to model spatial objects whose attributes have a certain degree of membership. Spatial and non spatial attributes can support the information related to vague regions. Fuzzy Logic is crucial in modeling vague regions since their borders are not sharply defined. In the application presented here, Fuzzy Logic helps us to determine the degree of membership of the different towns to each risk zone and then can determine which evacuation road could be the most suitable one. Therefore, Association Rules obtained from FDC will assist on the decision making process, for instance, when it is necessary to choose among different available evacuation routes, it has to be considered quality measures such as confidence and support that association rules provide.