Articulating morphological and semantic data in heritage modelling. A conceptual multilayer framework for AI-assisted decision support in architectural heritage

Abidi, Zayneb;Barouni, Khouloud;Claeys, Damien;Naifer, Zaineb
(2026) Proceedings of the 7th International Symposium on Human, Society and Sustainable Development Research — ISBN: [978-625-92747-8-2], 945-961, published

Files

125_260713_Heritage-modeling-MEP.pdf
  • Open Access
  • Adobe PDF
  • 494.67 KB
Download

Details

Authors
Abstract
(en) Current digital practices allow a faithful acquisition of the volumetries of heritage objects, but they struggle to integrate the spatial and cultural logics that generated the architecture. In the era of artificial intelligence (AI), an epistemological step back appears necessary to rethink heritage modelling methods when they combine scattered data arising, simultaneously, from the quantitative reproduction of geometries, the qualitative integration of the design process, and semantic appropriation. Building on a systemic approach to architectural heritage, this article proposes a conceptual and methodological framework aimed at producing hierarchised knowledge from heterogeneous data. The data are organised within a multilayer matrix: (1) a morphological layer, describing the physical, geometric and configurational characteristics of the built fabric; (2) a semantic layer, integrating historical knowledge, uses, cultural values, and architectural rules; (3) a relational layer, supported by AI, that articulates morphological and semantic information through explainable morpho-semantic relationships. This organisation constitutes a structured heritage dataset Embedded within a reproducible and systematically structured processing pipeline, spanning from data acquisition to knowledge generation. Within this framework, AI is not envisaged as a substitute for historical expertise, but as a transversal reasoning operator capable of identifying inter-layer regularities, making correlations explicit, and inferring interpretable architectural rules opening the way to interpretive and predictive modelling. Rather than presenting an operational decision-support system, the proposed framework establishes the conceptual and methodological foundations for future AI-assisted heritage applications. The proposed approach thus models heritage documentation as a dynamic decisionsupport system, contributing to the analysis of historical data, to digital reverse engineering, and to the methodological foundations of conservation and valorisation strategies.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Abidi, Z., Barouni, K., Claeys, D., & Naifer, Z. (2026). Articulating morphological and semantic data in heritage modelling. A conceptual multilayer framework for AI-assisted decision support in architectural heritage. In Baskin, S., Sid, A. S. (ed.), Proceedings of the 7th International Symposium on Human, Society and Sustainable Development Research (pp. 945-961). Recent Academic Studies. https://hdl.handle.net/2078.5/279261