Historical place names: from archives to Linked Open Data

Place names and administrative boundaries are changing over time. The importance of historical place names and administrative/religious boundaries is widely recognized by scholars. In implementing a geographic names repository, several issues emerge, especially if the considered time range spans several centuries. Historical data get value if they can be put in their context, and this feature requires a solid data infrastructure. The pilot study leading to TGN relied on a database structure. The "ontological" approach and the LOD paradigm are offering even bigger advantages: interoperability and openness are the most relevant, because any information modeled using Semantic Web standards (like RDF and OWL) can be freely accessed and referenced by any web application. In addition, information is not bounded to be hosted on a single site/repository, but can be distributed everywhere on the Web. The project currently under way aims to make historical place names available according to the LOD paradigm. The first data sample has been the one used in the previously recalled pilot study. Even if the ontology conforms to the "golden rules" for Linked Open Data, it is not fully satisfactory, as there is little reference to shared ontologies. Therefore, a new version of the ontology, with greater emphasis on events as the cause of changes, is currently under development. Combining several available ontologies (including CIDOC CRM) data will be represented as a set of triples, as required by the underlying RDF model, and made publicly available as LOD, while a HTML5 application will support navigation, querying and rendering of data. The appropriate framework to support map interaction as well as the possibility of supporting some kind of "social" contribution (comments and gathering additional information) are currently under investigation.