Controlled Vocabulary
Glossary
This glossary sets out the controlled vocabulary used across the Memory Twin Framework. Each entry gives a formal one-sentence definition followed by an elaboration, and, where a term is governed by an existing standard, the authoritative source is cited. Definitions here are those applied in the framework's peer-reviewed publications and in its cultural heritage 3D digitisation practice.
Memory Twin
A Memory Twin is a cultural heritage digital record in which measured three-dimensional geometry is bound inseparably to its metadata, its paradata and the intangible cultural values, uses and testimonies that give the heritage asset its meaning.
The Memory Twin extends the Digital Twin by treating geometry as one layer among several rather than as the record itself. Where a Digital Twin answers what a monument looks like, a Memory Twin also answers how and why it was documented, what remained uncertain, which interpretive decisions were taken, and what the place or object means to the community that holds it. The concept was formalised through the UNESCO Chair on Digital Cultural Heritage at the Cyprus University of Technology and demonstrated at Villa Portelli in Malta and at Lambousa and Fikardou in Cyprus. Its documentation requirements draw directly on the transparency obligations of the London Charter and the Seville Principles, and on the quality parameters set out in VIGIE 2020/654.
Digital Twin
A Digital Twin is a computational replica of a physical asset that mirrors its measurable geometry, materials and, in engineering contexts, its live operational state through linked sensor data.
Originating in manufacturing and infrastructure management, the Digital Twin was conceived to support simulation, monitoring and predictive maintenance. Applied to cultural heritage it delivers accurate survey geometry, point clouds, meshes, orthophotography, and supports condition monitoring and structural analysis. Its limitation in heritage contexts is epistemic rather than technical: the model records state without recording evidence, so the reasoning, uncertainty and cultural meaning behind the record are absent. VIGIE 2020/654 addresses the measurable half of this problem by setting quality and complexity parameters for 3D digitisation; the Memory Twin addresses the remainder by attaching paradata and intangible values to the same asset.
Paradata
Paradata is the documented record of the processes, methods, instruments, assumptions and interpretive decisions through which a digital heritage record was produced, together with the uncertainties that remain unresolved.
Paradata is distinct from data, which is the record itself, and from metadata, which describes the resulting resource. It answers why a record looks as it does: which instrument was used and at what resolution, what site conditions constrained capture, which processing choices were made, which reconstructed elements are hypothesis rather than measurement, and what confidence attaches to each. The London Charter established paradata as an ethical requirement for computer-based visualisation, holding that the intellectual transparency of a visualisation is a precondition for its scholarly use; the Seville Principles restate this for virtual archaeology. Within the Memory Twin Framework paradata is machine-readable and travels with the model, so that reasoning survives the dispersal of the survey team.
Metadata
Metadata is structured descriptive information about a digital resource, its identity, provenance, rights, format, spatial reference and relationships to other resources, that enables the resource to be found, understood, managed and preserved.
In cultural heritage practice metadata is normally expressed through recognised schemas such as Dublin Core, LIDO, EDM or CIDOC-CRM alignments, and is anchored by persistent identifiers. Metadata makes an asset discoverable and legally usable; it does not, on its own, make it verifiable. A file may carry exemplary metadata and still leave a researcher unable to judge whether a wall in the model was surveyed or inferred. The Memory Twin Framework therefore treats metadata and paradata as a paired obligation, with VIGIE 2020/654 providing the reference parameters for what metadata a quality 3D heritage record should carry.
HHBIM
Heritage or Historic Building Information Modelling (HHBIM) is the structuring of a historic building as a semantically rich parametric model in which each element carries its construction phase, material, condition, intervention history and associated documentation.
HHBIM adapts BIM methods, designed for new construction, to fabric that is irregular, stratified and only partially knowable. Elements are derived from survey rather than from design intent, which makes explicit treatment of tolerance and hypothesis unavoidable. The Memory Twin Framework adopts HHBIM as the natural carrier for built heritage, extending it in two directions: paradata is attached at element level so that evidence and uncertainty travel with the geometry, and intangible values such as oral testimony and ritual use are bound to the same objects. Villa Portelli in Malta is the framework's principal HHBIM demonstration, undertaken ahead of adaptive reuse.
Intangible Cultural Heritage
Intangible Cultural Heritage comprises the practices, representations, expressions, knowledge and skills, together with the instruments, objects, artefacts and cultural spaces associated with them, that communities recognise as part of their cultural heritage.
The definition follows the UNESCO Convention for the Safeguarding of the Intangible Cultural Heritage (2003), which covers oral traditions, performing arts, social practices, rituals, festive events, knowledge concerning nature, and traditional craftsmanship. Intangible heritage is transmitted between generations and is continually recreated, which makes it resistant to fixed documentation. The Memory Twin Framework treats it as a first-class layer rather than as commentary: testimony, use and association are encoded against the same model elements as the measured fabric, so that conservation and adaptive reuse decisions taken through the model are informed by meaning as well as material.
FAIR Principles
The FAIR Principles require that research data be Findable, Accessible, Interoperable and Reusable, for machines as well as for people.
Published in 2016, FAIR specifies that data carry globally unique persistent identifiers and rich metadata, that they be retrievable through open and standard protocols, that they use shared vocabularies and formal knowledge representation, and that they be released with clear provenance and licensing. Applied to cultural heritage, FAIR is what turns a 3D record from a downloadable file into a citable scientific resource: DOIs make assets findable, open formats make them accessible, CIDOC-CRM alignment makes them interoperable, and metadata with paradata makes them genuinely reusable. VIGIE 2020/654 embeds FAIR expectations in its recommendations for European 3D heritage digitisation.
CARE Principles
The CARE Principles for Indigenous Data Governance require that data serve Collective Benefit, that communities retain Authority to Control, that those handling data act with Responsibility, and that the whole be governed by Ethics.
CARE was formulated by the Global Indigenous Data Alliance as a people-and-purpose counterweight to the technically-oriented FAIR principles. It asserts that openness alone can reproduce extraction: a dataset may be perfectly interoperable and still be published without the consent, benefit or authority of the community whose memory it holds. In the Memory Twin Framework, CARE governs who decides what is recorded, what is published, what remains restricted, and how intangible practice is represented. FAIR and CARE are applied together, the first to the record and the second to the relationship, so that scientific rigour and community authority are not traded against each other.
Level of Detail (LOD)
Level of Detail is the declared degree of geometric resolution, completeness and semantic richness at which a heritage asset has been recorded or modelled.
LOD is used in two related senses: the resolution of captured geometry, expressed as point spacing or surface accuracy, and the modelling grade of a BIM or HHBIM element, expressed as its development stage. Both are declarations of fitness for purpose; a record adequate for interpretation may be inadequate for structural analysis or for reconstruction after loss. VIGIE 2020/654 provides the European reference for complexity and quality parameters in 3D digitisation of tangible cultural heritage, and the Memory Twin Framework requires LOD to be stated explicitly as paradata rather than inferred, so that later users can judge whether a record supports the use they intend.
Semantic Interoperability
Semantic interoperability is the capacity of independent systems to exchange information in such a way that the meaning of that information is preserved and correctly interpreted by the receiving system.
Technical interoperability moves bytes between systems; semantic interoperability ensures that a field labelled date means the same thing on both sides, whether it refers to date of construction, of survey, or of acquisition. In cultural heritage this is achieved through shared ontologies, controlled vocabularies and authority files, with CIDOC-CRM as the principal reference model and thesauri such as the Getty AAT supplying terms. The Memory Twin Framework depends on semantic interoperability because a Memory Twin is by construction an aggregation of heterogeneous sources, survey, archive, testimony, condition records, and because European infrastructures expect heritage records to federate rather than to sit in isolation.
CIDOC-CRM
CIDOC-CRM is the ISO 21127 conceptual reference model for cultural heritage information, providing a formal, event-centred ontology for describing the entities, actors, places, times and relationships documented by museums, archives and heritage bodies.
Rather than prescribing fields, CIDOC-CRM models the events through which things come to be: a building is constructed, modified, surveyed, damaged and reused, and each event links actors, places, times and objects. This event-centred structure suits heritage particularly well because it records the history of the record alongside the history of the thing. Within the Memory Twin Framework, CIDOC-CRM and its extensions provide the semantic backbone that binds geometry, metadata, paradata and intangible values into a single queryable graph, and they are the mechanism by which framework outputs satisfy the interoperability requirement of the FAIR principles.
Knowledge Graph
A knowledge graph is a machine-readable network of entities and typed relationships, expressed against a formal ontology, that allows information from heterogeneous sources to be integrated and queried as a single structure.
In heritage practice a knowledge graph links a building to its construction events, its architects, its archival documents, its survey campaigns, its condition reports and the communities associated with it, each node carrying a persistent identifier. Expressed in RDF and aligned to CIDOC-CRM, such a graph can be queried across institutions and enriched without restructuring. The Memory Twin Framework treats the knowledge graph as the operational form of the Memory Twin: the 3D model is the geometric expression, while the graph carries the reasoning, provenance and meaning, keeping paradata addressable at element level rather than buried in a project report.
Digital Repatriation
Digital repatriation is the return of high-quality digital records of cultural heritage, and of the rights to interpret and control them, to the communities and places of origin from which the material heritage was removed.
Digital repatriation does not substitute for the physical return of objects and should not be presented as doing so. Its value is access and authority: a community regains the ability to study, teach with, reinterpret and reuse representations of its own heritage, and to govern how those representations circulate. Done well, it is CARE applied to 3D data; done poorly, it becomes another act of extraction in which an institution retains the master files and the narrative. The Memory Twin Framework treats repatriated records as needing full paradata, so that receiving communities can assess what was captured, at what quality and under what assumptions.
Adaptive Reuse
Adaptive reuse is the conversion of a historic building or site to a new function while retaining the significance, fabric and cultural values that make it heritage.
Adaptive reuse is often the only realistic route to the long-term survival of a building whose original function has lapsed, but it is also the point at which significance is most easily lost, because intervention decisions are taken quickly and on incomplete evidence. A Memory Twin supports the process by making evidence, uncertainty and intangible value available at the moment of decision: which elements are original, which are later interventions, which are hypothesis, and which carry community meaning that survey alone would not register. Villa Portelli in Malta is documented within the framework precisely as an adaptive reuse case of this kind.