After the fire that destroyed part of the cathedral Notre-Dame de Paris, a working group specialized in digital data coordinated a scientific project that would allow the management of all the digital data produced by scientific research activi-ties along with restoration operations. The ERC advanced grant “nDame Heritage” project combines digital humanities with computer science and artificial intelli-gence to create a collaborative knowledge system that analyzes multiple views from different experts on the same cultural heritage objects. In this work, we designed a hybrid artificial intelligence workflow based on both knowledge graphs and deep learning models for semantic segmentation of 2D images. We show that this hybrid approach can help experts to process, integrate, and enrich Notre-Dame’s data.
Versioning Virtual Reconstruction Hypotheses: Revealing Counterfactual Trajectories of the Fallen Voussoirs of Notre-Dame de Paris using Reasoning and 2D/3D Visualization
Virtual reconstruction should move beyond merely presenting 3D models by documenting the scientific context and reasoning underlying the reconstruction process. For instance, the collapsed arch in the nave of Notre-Dame de Paris serves as a case study to make explicit the reconstruction argumentation encapsulated in relation to the spatial configuration of the arch and the voussoirs. The experiment is twofold: (1) setting up of the 3D dataset where the hypotheses are modeled as versions using logic programming, and (2) evaluating the scientific narrative of reconstruction through both a custom 2D-3D visualization and competency questions on the enriched 3D data. Formalization, reasoning, and visualization are combined to explore the nonlinear scientific hypotheses and narrative of the reconstruction. The results explicitly show both the factual information on the physical and digital objects, as well as the counterfactual propositions allowing the reasoning at play in the reconstruction. The hypotheses are visualized as counterfactual trajectories creating an open dynamic visualization that makes possible the spatialized querying of conflicting interpretations and embedded memory in place.