Um framework para data warehouse espacial.

Abstract

DataWarehouse (DW) and OLAP tools (Online Analytical Processing) are two echnologies widely used to help managers in decision support process. Recent researches attempted to integrate spatial data into Data Warehouse environment. Most of the work related to spatial DW focus on the definition of new data models and implementation of new OLAP tools. The need for new OLAP tools is justifiable, since traditional OLAP tools are not prepared to deal with spatial data, so a new category of OLAP tool with special features comes up. This new category is referred by the term Spatial OLAP (SOLAP). Spatial OLAP aims to integrate features of OLAP tools with features traditionally found in the Geographic Information Systems (GIS). Despite the large number of SOLAP tools found in literature, we believe that OLAP-GIS integration is not fully satisfactory for the following reason: there is no tool that combines the use of spatial and multidimensional operators and is intended to end users. Therefore, this work proposes a new SOLAP tool called Mapwarehouse. Mapwarehouse lets end-users define their own Spatial DataWarehouse schemas and query Spatial Data Warehouses created from these schemes. For this, the tool has a graphical language for defining schemas and queries, implemented as a set of graphical interfaces on a high level of abstraction. The proposed framework follows the framework philosophy, which allows programmers to extend it and use it in different technologies of Database Managment Systemas (DBMS). Aiming to validate it, the framework was extended to Oracle 10g and PostgreSQL. Furthermore, we developed two case studies: one for the analysis of the agricultural distribution of state of Paraíba and another for the analysis of the care provided by the SAMU at Aracaju city, Sergipe.