Tiago Leão, November 2011

Caracterização Espacial utilizando Indução Orientada aos Atributos no SOLAP+, João Moura Pires (superv.), Universidade Nova de Lisboa, November 2011.
Keywords: SOLAP; data mining; induction; generalization; hierarchies; spatial characterization

Abstract: The prototype SOLAP+, developed on the direction of João Moura Pires, follows the concept SOLAP presented by Bédard et al. combining the functions of support systems for decision OLAP with the capacities of geographical information systems (GIS). With the work undertaken by Rose Martins (2006), Marlene Vitorino and Rodolfo Caldeira (2008), Ruben Jorge (2009) and Ricardo Silva (2010) was developed a generic approach SOLAP that culminated with the implementation of a prototype that, currently, already presents a few features from the point of view of analysis of data and representation of that data using thematic maps and/or graphics. To assist in this task of analysis, we integrate mechanisms to discover relationships between the data and present these relations on the map. This analysis will enable the user to take conclusions by means of information that is not explicit in the data present in the database. Thus, the aim of this dissertation is to incorporate, with the multidimensional model followed, two induction techniques: spatial dominant generalization and non-spatial dominant generalization. With the results of applying these techniques, we semantically characterize the spatial objects in the map.