Implementação de uma arquitetura para minerar regras de associação espaciais a partir de dados espaciais fuzzy

Abstract

The use and importance of spatial data have been growing more and more. With the assimilation of the real world, many spatial phenomena are modeled and represented by fuzzy spatial objects, which have blurred interiors, uncertain boundaries and/or inexact locations. Inferring rules and finding patterns is something that human beings have always done, with the advancement of Artificial Intelligence and the representation of a spatial object, this work proposes an implementation of an architecture capable of mining spatial association rules from fuzzy spatial data. With user input as a parameter, the proposed architecture discovers patterns of spatial association between spatial objects of interest to users (e.g. tourist attractions) and such fuzzy spatial features (e.g. sanitary conditions of restaurants, number of reviews and price of accommodations). Furthermore, this work presents a case study based on real datasets that shows the applicability of the developed method.