Towards vague geographic data warehouses

Abstract

Currently, geographic data warehouses provide a means of carrying out spatial analysis together with agile and flexible multidimensional analytical queries over huge volumes of data. However, they do not enable the representation and neither the analysis over real world phenomena that have uncertain locations or vague boundaries, which are denoted by vague spatial objects. In this paper, we introduce the vague geographic data warehouse (vGDW) and its spatially-enabled components at the logical level: attributes, measures, dimensions, hierarchies and queries. We base the vGDW on exact models to represent vague spatial objects. In addition, we combine the fuzzy model with the exact model in relational vGDW to improve the expressiveness of the queries. Finally, a case study is presented to validate our contributions. © 2012 Springer-Verlag.