SEMANTIC AUGMENTATION OF GEOSPATIAL CONCEPTS: THE MULTI-VIEW AUGMENTED CONCEPT TO IMPROVE SEMANTIC INTEROPERABILITY BETWEEN MULTIPLES GEOSPATIAL DATABASES

摘要

Semantic interoperability is a key issue for the meaningful sharing of geospatial data between multiples geospatial databases. It requires the establishment of semantic mappings between concepts databases' ontologies. Semantic mappings can be discovered only when semantics is explicit. However, existing concepts' definitions are not always sufficient to represent the semantic richness of geospatial concepts. In addition, semantics may be implicit, refraining from using it during semantic mapping process. This paper, proposes a new representation for geospatial concepts, called Multi-View Augmented Concept (MVAC), which takes into account these drawbacks. We propose a method to generate a MVAC, based on: (1) extraction of the different views of a concept that are valid in different contexts, and (2) augmentation of a concept with implicit dependencies between its features based on rule mining theory. We believe that the proposed approach will play an important role to improve the quality of the semantic interoperability between multiple geospatial databases since it takes into account the implicit semantic relations between different concepts.