Ontology and Spatial Planning

摘要

In last decades a key problem in adopting technologies in planning process was a chronic lack of data. But in recent times, such problem was inverted due to the overabundance of data produced in different periods, with various purposes, at multiple scales and with different cognitive models. This situation generated three types of barriers to data interoperability: bureaucratic, technological, semantic. While the first two issues have been solved taking various initiatives, the last one could be solved using ontologies. Concepts are the cornerstone of the ontology, but it is not easy to define a concept without any ambiguity, discordance or vagueness. A concept can be clear or not; ambiguity occurs when a concept is not much clear; while discordance arises when an agreement is missing. If the concept definition can present some incoherence, the broad boundaries model can be useful in Ontology representation. This model is an extension of the 9-intersection model used for the topological relationship among geographical objects. The model with broad boundaries deals with uncertainty in spatial data taking into account ill defined aspects. This model is based on the definitions of inner and broad boundaries. Using this model in Ontology field, the inner boundary is the edge of the part of a concept without doubts and the broad boundary is the grey zone, with a certain level of uncertainty, useful to represent ambiguity, discordance and vagueness. Topology rules represent the relationship among concepts. If two concepts are identical, the "equal" rule can be used; if they share some parts, the "overlap" rule is suitable. If two concepts are completely different, the "disjoint" rule can be applied. If a concept is a subset of another, there are several rules which can help us ("covers", "covered by", "contains" and "inside"). In case all concepts are clear, these relationships can be modelled using the 9-intersection model. The way to define the part of concept included inside the inner boundary and the other one included in the broad boundary can be achieved using rough set theory. All the aspects of a concept classified in the same way represent the indiscernible part of the concept and are included inside lower approximation (inner boundary). The remaining part represents an uncertainty zone and it falls within the upper approximation (outer boundary). The measure of the degree of uncertainty inside the upper approximation can be modelled using fuzzy set theory. This approach has been tested with several concepts particularly suitable to verify the hypothesis.