Towards the extraction of semantics from incomplete archaeological records

The Archaeological Geographical Information System of Verona (SITAVR) has been developed since 2011, starting from an existing and well-consolidated system, the Archaeological Geographical Information System of Rome (SITAR). The main objective of both projects is collecting information about the archaeological findings regarding the two Italian urban centres with the aim to support a complete archaeological analysis and allow for easy data reuse and dissemination. Data collected in both projects come not only from current excavation campaigns, but they have also been extracted from very old documents describing excavations performed during the 18th century. In this last case, many details about the location and the time of existence of findings are not known. The SITAVR project has lead to the definition of a domain model for archaeological data, called Star, and now it is focused on the mapping between this model and other international standards, such as the CIDOC CRM and its extension CIDOC-CRMarcheo. This paper presents a general approach for guiding the mapping of a dataset regarding ancient archaeological data towards CIDOC-CRMarcheo, especially for what concerns to spatial and temporal properties that are only partially specified.