Fundamentals of spatial data warehousing for geographic knowledge discovery

Recent years have witnessed major changes in the geographic information (GI) market, from interoperable technological offerings to national spatial data infrastructures, Web-mapping services, and mobile applications. The arrival of new major players such as Google, Microsoft, Nokia, and TomTom, for instance, has created tremendous new opportunities and geographic data have become ubiquitous. Thousands of systems are geo-enabled every week, including data warehouses. As a special type of databases, a data warehouse aims at providing organizations with an integrated, homogeneous view of data covering a significant period in order to facilitate decision making. Such a view typically involves data about geographic, administrative, or political places, regions, or networks organized in hierarchies. Data warehouses are separated from transactional databases and are structured to facilitate data analysis. They are built with a relational, object-oriented, multidimensional, or hybrid paradigm although it is with the two latter that they bring the most benefits. Data warehouses are designed as a piece of the overall technological framework of the organization and they are implemented according to very diverse architectures responding to differing users’ contexts. In fact, the evolution of spatial data warehouses fits within the general trends of mainstream information technology (IT).