ARXIVE Data Alignment Results I (D4.1)

This document constitutes the Methodology and Data Provenance Annex to ARXIVE Deliverable D4.1, Data Alignment Results I and presents the semantic alignment methodology developed under Task 4.1, together with its first implementation and validation results. The methodology aims to support the transformation of heterogeneous cultural heritage data into interoperable, semantically structured representations, while preserving source meaning, provenance and traceability. The first implementation focuses on the National Library of Greece (NLG) conservation data, using two complementary domains: Prints & Drawings and Books. The methodology was initially developed and validated using the documented Prints & Drawings source fields and structures and subsequently reused and extended for Books. This approach enabled the identification of reusable semantic patterns alongside domain-specific requirements and demonstrated how a common mapping framework can accommodate differences in source structure and terminology. CIDOC CRM provides the primary semantic foundation, complemented by relevant extensions and models where required by the source data or interoperability context. The methodology combines source analysis, field-level semantic mapping, reusable graph patterns, controlledvocabulary alignment and iterative validation. Vocabulary alignment draws on established authorities, including Getty AAT, Ligatus Language of Bindings, RBMS Controlled Vocabularies and the Thesaurus for Graphic Materials, while preserving locally defined concepts where external equivalence cannot be established. The resulting outputs include validated mapping registers, machine-readable vocabulary resources, an exploratory cross-domain semantic baseline, a reference RDF transformation (A06) and a reproducible RML transformation package (A08). Representative NLG records demonstrate the progression from source data through semantic mapping to RDF generation. Validation combines semantic consistency checks, graph-level examination, SHACL validation, competency queries and a comparative assessment of the independently developed A06 and A08 transformations. This comparison identifies both shared modelling outcomes and differences in semantic representation, without treating structural conformance or reproducibility as evidence of full semantic equivalence. The Annex also documents confirmed alignments, conditional reuse decisions, remaining limitations and requirements for subsequent integration. The results establish a first validated, pilot-level semantic baseline for downstream ARXIVE knowledge graph integration, data quality management and repository synchronisation workflows. While the methodology has been implemented and tested within the examined NLG domains, its broader cross-domain transferability and operational integration remain subject to further validation. Subject to the availability of sufficient archaeological source material, the next stage is intended to evaluate the methodology in the MMP pilot, assessing its transferability across institutional and data contexts and identifying additional modelling requirements. The accompanying artifacts provide a documented basis for this continued assessment and for subsequent implementation and refinement within the relevant ARXIVE tasks, particularly T3.1, T3.2 and T4.2, in the context of the European Collaborative Cloud for Cultural Heritage (ECCCH).