Spatial ontologies for architectural heritage

Informatics and artificial intelligence have generated new requirements for digital archiving, information, and documentation. Semantic interoperability has become fundamental for the management and sharing of information. The constraints to data interpretation enable both database interoperability, for data and schemas sharing and reuse, and information retrieval in large datasets. Another challenging issue is the exploitation of automated reasoning possibilities. The solution is the use of domain ontologies as a reference for data modelling in information systems. The architectural heritage (AH) domain is considered in this thesis. The documentation in this field, particularly complex and multifaceted, is well-known to be critical for the preservation, knowledge, and promotion of the monuments. For these reasons, digital inventories, also exploiting standards and new semantic technologies, are developed by international organisations (Getty Institute, ONU, European Union). Geometric and geographic information is essential part of a monument. It is composed by a number of aspects (spatial, topological, and mereological relations; accuracy; multi-scale representation; time; etc.). Currently, geomatics permits the obtaining of very accurate and dense 3D models (possibly enriched with textures) and derived products, in both raster and vector format. Many standards were published for the geographic field or in the cultural heritage domain. However, the first ones are limited in the foreseen representation scales (the maximum is achieved by OGC CityGML), and the semantic values do not consider the full semantic richness of AH. The second ones (especially the core ontology CIDOC – CRM, the Conceptual Reference Model of the Documentation Commettee of the International Council of Museums) were employed to document museums’ objects. Even if it was recently extended to standing buildings and a spatial extension was included, the integration of complex 3D models has not yet been achieved. In this thesis, the aspects (especially spatial issues) to consider in the documentation of monuments are analysed. In the light of them, the OGC CityGML is extended for the management of AH complexity. An approach ‘from the landscape to the detail’ is used, for considering the monument in a wider system, which is essential for analysis and reasoning about such complex objects. An implementation test is conducted on a case study, preferring open source applications.