Alignment-based semantic translation of geospatial data

Recent years are characterized by a renewed interest in semantic technologies. Growing number of data providers and users facilitate production and consumption of semantically-annotated resources. Unfortunately, this process does not go hand-in-hand with “stabilization” (not to talk about standard-ization) of a set of broadly accepted ontologies. As a matter of fact, the number of ontologies being proposed/developed, often “competing” with each other within the same domain, systematically increases. This, in consequence, stimulates work devoted to, various forms of, ontology “reconciliation”, such as: establishing alignments and/or ontology merging. In this context, in the Internet of Things (IoT), use of semantic technologies materializes, among others, in the han-dling of messages exchanged between artifacts that constitute IoT ecosystems. Here, some existing platforms (e.g. Open-IoT, UniversAAL or VITAL) use serializations of RDF to transmit messages, others (e.g. FIWARE or oneM2M) support ontologies, even if they don't use RDF. Thus, one of important unresolved issues remains: how to facilitate understanding of messages exchanged between artifacts founded on different ontologies (i.e. establish semantic interoperability). In this paper, we focus on semantic translations, based on ontology alignments. We discuss the problem of identifying and representing alignments (using geospatial data as the use case example), so that they can be further used in the translation process.