An Architecture and a Metamodel for Processing Analytic and Geographic Multilevel Queries

Analytic and geographic decision support queries are characterized by performing spatial analysis on data aggregated at different levels of detail and are typically large and complex to be written from scratch (manually) by a non-specialist user. Considering that decision support tools are in evidence and many databases (DB) have information with some geographic reference, nonspecialist users of these DB need analytic and spatial decision support tools to better analyze DB information. In this context, Spatial On-Line Analytical Processing (SOLAP) tools have received a lot of attention. Nevertheless, as there is no de jure standard language for OLAP yet, like standard SQL is to Relational Database Management Systems (RDBMS), these tools are dependent on specific OLAP languages, Application Programming Interface (API) and servers. In order to propose an alternative to this problem, this paper presents an Analytic and Geographic Information Service (AGIS). Our proposal is based on open and extensible standards (i.e. it is not based on an OLAP server) to offer a service that provides multilevel analytic functions that aim to enrich the set of functionalities of Geographic Information Systems (GIS). In order to do this, AGIS 1) abstracts the complexity of generating analytic and geographic decision support queries and 2) is independent of hardware and software platform. To satisfy these goals, a three-tiered architecture and a metamodel were defined and implemented. As a proof of concept, a study case to analyze the electrical energy situation in Brazil was implemented using our proposal.