IN-PERSON: Prague – Section B

Abstract

Architectural heritage is the embodiment of the peoples, cultures, methods, and situations that shaped it and imbued it with meaning. All the clues it embodies about its timeline are often captured in different formats and have to be consolidated for use by present and future researchers. Computational design in architectural heritage emerges as a new field that goes beyond digital representation tools. It is instrumental in producing multiple layers of information on the design of a building, adding to its value as a historical document. It also aids in packaging and disseminating compiled information such that it can be readily added to by others. This study presents and reflects on the challenges of data processing for a sample of Seljuk-era monuments in Turkey. The data came from various sources and an interdisciplinary team focused on drawing meaning from it using digital techniques such as semantic segmentation with deep learning and information modeling. However, literature and previous surveys were insufficient in providing information on the construction methods, structural systems and craft techniques. These are necessary for historical building information models which require detailed input for new ontologies particular to heritage. Such information may have also allowed the sample monuments to be better contextualized among others in the region. Any heritage asset (and the heritage field itself) is a living and dynamic well of information. Documentation of heritage should encapsulate this and allow information to be built up as an album, rather than remain a static snapshot. The study discusses these issues in detail and offers suggestions about creating editable, open-source digital libraries to reflect the documentary value of any heritage asset and meeting the need to allow information to be built up as an album, rather than remain a frozen snapshot.