Semantic relations in ontology mediated linguistic data integration
Alexis Dimitriadis, Adam Saulwick, Menzo Windhouwer · UvA-DARE (University of Amsterdam) · 2005
In developing a system for integrated access to diverse typological databases, we rely on the formulation of general principles by which linguistic concepts and database fields and their relationships can be appropriately modeled. By modeling these entities and relationships, representing them in an ontology and linking them to a merged hierarchical data structure, heterogeneous perspectives on linguistic phenomena can be managed in a unified way without distorting the source data. In this paper we enumerate these principles and present our approach to data integration with a range of specific examples from component databases, which illustrate the relationship types and integration methods employed.