A Semantic Approach to Constraint-Based Reasoning in Geographical Domains

Various models have been developed to manage geographic data but most of them integrate heterogeneous techniques to support knowledge representation and reasoning. This is far from optimal because it requires mapping data between different representation formats; moreover, as it fragments knowledge, it limits the possibility to use complete information about the problem to be solved for the execution of inferences. In order to address this issue, we adopt a unified approach, in which we use Semantic Web techniques to manage both knowledge representation and reasoning rules with particular attention to constraint verification that is central to several geographic reasoning tasks. Our model exploits an ontological description of spatial constraints which supports the specification of their properties, facilitating the automated selection of the relevant ones to be applied to a given problem. The model supports different types of inferences, such as checking the compliance of a given geographical area to a set of constraints, or suggesting a suitable aggregation of land patches that satisfy them. We test our model by applying it to the management of Ecological Networks, which describe the structure of existing real ecosystems and help planning their expansion, conservation and improvement by introducing constraints on land use.