Knowledge Acquisition of the Architectural Heritage with Classification and Fromalisation of Semantic Relationships: Development of Ontology Driven Knowledge-Bases

Document Type : Research Paper

Author

Assistant Professor and Member of Academic Board, Faculty of Architecture, University College of Fine Arts, University of Tehran

Abstract

Knowledge-based study of architecture comprises from understanding physical and abstract entities of architectural value, their categorization and defining semantic relationships between entities based on their specific or generic properties. In current decade advanced digital technologies of information and communication and intelligent systems tries to conceptualize phenomena and realize the semantic relationships between objects and concepts of the target domain in computer understandable languages. One of the important knowledge acquisition systems for knowledge share and reuse is Ontology. Ontology was a philosophical theory and now has new application in computer science for “specification of conceptualization (Gruber, 1993)” of different domains. Major concerns of ontologic systems is rigorous analysis of entities of a domain, their classification, and define of semantic relationship links between entities by using object oriented languages such as RDF or OWL. For design and development of an advanced ontologic knowledge acquisition system for Persian architecture we have used a tool with user friendly interface named Protégé to be able to avoid programming code of the languages such as RDF. At first we developed a terminology schema and studied different vocabularies and their meaning, by reviewing main upper level resources of dictionaries, encyclopedias of Islamic architecture or resources containing technical architectural terms. We have restored the gathered vocabularies in a terminology ontology schema mapped from General Ontology for Linguistic Descriptions (GOLD). Later we gave structure to the terminology by defining classes and subclasses mapped from upper level ontologies such as SUMO or inspired from Art and Architecture Thesaurus ®. For example a mosque in our schema is subcategory of ceremonial buildings, as a single work buildings, and subclass of the built environment. Definition, pronunciation, synonym, etymology of mosque was also provided as terminological properties. Lastly we started to model the semantic relationships in the ontology schema. We mapped the relationships from metadata standards such as Core Data Index to exclusively describe a building by defining physical properties such as names and resources (data), location, components, objects, materials, agents or abstract attributes such as activities, styles and periods, physical attributes, etc. These properties take direct values as defined properties (such as an exact name of an architect of a building) or takes instances connected to set of detailed properties (such as location instance that connect set of values such as address, world coordinates, and localities of a building). Conclusively the ontology-based knowledge acquisition system developed by this research was used to document around 50 cases of buildings in Iran using descriptive core inventories. The paper demonstrated the instantiation of the ontology by one of the most important samples of Persian architecture: The Jami Mosque of Isfahan. Finally the research envisions to open the ontology-based knowledge model developed for Persian architecture online by web application server or web framework application tools to be able to provide HTML pages for the end users from the structured knowledge in normal browsers and to enrich accessibility to data of Persian architecture by advanced abilities of the system for query and searches.

Keywords