Toward an Architecture for Enhancing Semantic Interoperability Based on Enrichment of Geospatial Data Semantics

Toward an Architecture for Enhancing Semantic Interoperability Based on Enrichment of Geospatial Data Semantics

Mohamed Bakillah (Laval University, Canada) and Mir Abolfazl Mostafavi (Laval University, Canada)
DOI: 10.4018/978-1-4666-0327-1.ch003
OnDemand PDF Download:
No Current Special Offers


Semantic interoperability is needed to support meaningful data exchanges in distributed environments such as ad hoc networks of geospatial databases and geospatial web services. Even with the increasing popularity of ontologies to capture semantics, semantics of geospatial data are often too weak to support meaningful exchanges. In this chapter, the authors argue that semantically weak geospatial data can be enriched to enhance semantic interoperability. They propose a conceptual architecture designed to support enhanced semantic interoperability in dynamic networks that focuses on semantic enrichment. The proposed conceptual architecture includes a coalition management module, an ontology enrichment module, and a semantic mapping module; the modules perform different types of semantic enrichment and can support various semantic interoperability tasks. Within the different enrichment methods, the authors explain the role of global ontologies, arguing that they play a key role in a semantic interoperability framework. Finally, the authors illustrate with an application example the possibilities of such architecture.
Chapter Preview


In GIScience, ontologies are employed to capture the concepts and meanings that are considered valid to describe the geospatial domain (Fonseca, Egenhofer, Agouris, & Câmara, 2002a; Bittner & Smith, 2003; Brodeur, 2004). According to a review by Agarwal (2005) on the role of ontologies in GIScience, increasing interest is given to geographic ontologies. The idea of a universal ontology of geographic space has been studied in several researches (Bittner & Smith, 2003; Čeh, 2003; Kuhn, 2003; Grenon & Smith, 2004). A universal ontology of geographic space would certainly play an invaluable role, given that in GIScience, reference ontologies are also considered as a standard that supports translation between different geographic information systems or negotiation of meanings between communities that use different vocabularies (Agarwal, 2005).

However, there is currently no comprehensive, universal geographic ontology (Timpf, 2002; Agarwal, 2005), and this is partly due to the fact that GIScience is evolving discipline that draws from several other disciplines. Nevertheless, some global ontologies, which capture concepts of a high and generic level of abstraction, independently of the domain of application (Guarino, 1998; Kashyap & Sheth, 1998; Brodeur, 2004), are certainly useful for a variety of tasks, including tasks related to achieving semantic interoperability. For example, OpenCyc, which is part of the Cyc project that aims at creating a comprehensive ontology and knowledge base of common sense knowledge to support human-like reasoning of Artificial Intelligence (AI) applications, contains spatial concepts related to direction and orientation relations, relative position of objects, mereological relations, etc.

ISO TC204, document N271 provides a currently well-known definition of interoperability: interoperability is the ability of systems to provide services to and accept services from other systems and to use the services so exchanged to enable them to operate effectively together.” Semantic interoperability is also defined as the “knowledge-level interoperability that provides cooperating businesses with the ability to bridge semantic conflicts arising from differences in implicit meanings, perspectives, and assumptions, thus creating a semantically compatible environment based on the agreed concepts…” (Park & Ram, 2004, p. 597). It is also analogous to human communication (Brodeur, Bédard, Edwards, & Moulin, 2003; Kuhn, 2005). In order to be established, semantic interoperability requires the semantics of geospatial data to be explicit and rich enough; this ensures that differences and similarities in concepts used by different systems can be detected with appropriate algorithms and/or by human experts.

Complete Chapter List

Search this Book: