CHAD-KG makes the process of digitisation itself part of cultural heritage knowledge. Rather than representing only objects, titles and collection metadata, it models paradata: the instruments, procedures, transformations and decisions through which digital surrogates are produced. Built through an application profile aligned with CIDOC-CRM, LRMoo, CRMdig and Getty AAT, and materialised by a reproducible RDF pipeline, the graph transforms digitisation from an invisible technical service into a documented epistemic chain. This is its iconic idea: a digital object cannot be critically understood without the history of how it was captured, processed and integrated. The project’s use as a metadata source for a digital twin demonstrates how provenance can become operational within exhibition and research environments. Its wider significance reaches beyond heritage informatics. CHAD-KG establishes that faithful representation depends less on visual resemblance than on accountable transformation; the digital twin becomes credible when its genealogy, not merely its appearance, remains queryable.
