SOUSA, André Gomes de; http://lattes.cnpq.br/6119022255104066; SOUSA, André Gomes de.
Resumo:
Spatial Data Warehouse can be characterized as a multidimensional database that
includes spatial dimensions and spatial measures. Depending on the context, spatial
information are essential to analyze of data in Decision support systems. For example,
dynamic aggregation of maps in diverse levels, combined with descriptive data and numerical correspondents, must facilitate the decision taking excessively. This research proposes and define formally a multidimensional data model that tightly integrates spatial and non-spatial data – spatial multidimensional model. This model, suitable for Spatial Data Warehouse, is described in its structural and managing aspects, treating of the
multidimensional modeling of the spatial and analytical data and spatial OLAP operations.
Moreover, is discussed the object-relational implementation of space multidimensional
schemas in OR DBMS that deals with spatial data, taking in consideration questions related to performance of queries that involve spatial aggregation. To guarantee the good performance of these queries, some enhancing techniques were proposed. The model and the ideas proposed were validated through the implementation of an archetype that uses analytical and spatial data established in a case study. The archetype was
constructed and tested using the Oracle 10g DBMS. An experimental evaluation evidenced
the efficiency of the proposed enhancing algorithms.