Documentation

Abstract

The documentation of the working steps, the decisions made in a reconstruction, the applied method, and the results form one of the cornerstones of scientific practice. Over the centuries, scientific publication established itself with fixed basic principles, such as verifiability of methods, objectivity, disclosure of sources, comprehensibility of argumentation, accessibility of results, accuracy, reliability, and uniformity [1]. In computer-aided, hypothetical 3D reconstruction of destroyed architecture, the application of the above basic principles faces an as yet unsolved challenge. The model creation process is rarely documented, and when it is, the documentation is usually not publicly available. The knowledge embedded in reconstructions, scientific interpretation, argumentation, and hypothesis, is in danger of being lost. Due to the lack of resources, diverse and rapidly developing software applications, modelling methods and types, no application-based method for documenting and publishing 3D models has been established. Three decades into the spread of computer-assisted 3D visualization in the research and mediation of cultural heritage, discussion of the question of what and how to document has intensified. The Internet as a publication venue seems to make sense to most. Web-based documentation requires technical infrastructures and services as well as defined scientific methods for comprehensible modeling and sustainable provision. The following chapter is dedicated to describing and clarifying these developments.

Guiding questions

• What guidelines and projects exist?

• Why is documentation important?

• What needs to be documented?

• How can one document and publish?

Basic terms

• Metadata and paradata

• Scientificity and traceability

• Findability and accessibility

• Interoperability and reusability

8.1 Introduction

The topic of documenting 3D reconstructions is examined below under four questions:

• 1. Why is documentation important?

• 2. What kind of research data are we dealing with at all?

• 3. What should be documented?

• 4. How should documentation be done?

The consideration of these aspects has two focal points: firstly, the documentation of information accompanying a 3D model and secondly, the documentation of the decision-making processes related to the reconstruction. Documentation should provide information about the circumstances surrounding the reconstruction and the digital model behind it as well as information about the actors, backgrounds, and technical conditions. Documentation ensures that the occasion, purpose, background, sources, and the decision-makers and others’ creative considerations for and against creating a reconstruction in a certain way, are comprehensible. Documentation is part of good scientific practice [2]. In every scientific discipline, documentation of research results and the processes behind them is required in order to make findings and justifications comprehensible and available to future generations as a basis for further research.

Documentation of the decision-making processes and the sources behind them can be highly important for one’s own reconstruction practice. Knowledge and effort have gone into reconstruction models. These models therefore have a certain value. If one wants to use and change these models again later, it is important to be able to reconstruct the decision-making processes and the sources used in order to resume or repeat the work with as little effort of one’s own as possible. Furthermore, such documentation can strengthen one’s own reputation. If reconstructions are visible to the public and questions about their credibility arise, then documentation can help to communicate the earlier decisions and thus reject any doubts raised. If, on the other hand, objections are justified, then documentation helps one to make changes in a reconstruction model more quickly, more precisely, and thus more economically. Especially in the commercial sector, such as TV documentaries, entertainment, or reconstructions for museums, this can be advantageous.

8.2 Why Documentation is Important

8.2.1 Scientific Significance

Digital models have become widely established as tools of mediation and research in the context of architectural and urban studies [3]. Especially the increasing dissemination of digital architectural reconstructions in exhibitions and research projects and the public funding that often accompanies them raises questions about the sustainability of the reconstructions presented and the knowledge embedded in them. It is necessary to ask where and how reconstruction models and 3D datasets can be permanently presented, secured, and accompanying information documented. Yet it is challenge to document decision-making processes and the underlying thought processes and interpretations in digital reconstructions in a publicly accessible, transparent, and comprehensible way. If this does not succeed, the knowledge and thus the potential scientific value of a reconstruction risks being lost. Simply saving the results is not sufficient for valid documentation. The publication of research results from digital 3D reconstructions in print media, television, exhibitions and the Internet, including pure 3D repositories such as Sketchfab.com, usually only represent the result of research processes along with rudimentary accompanying information, but not their genesis and the discourse conducted in the process. Moreover, the lack of documentation of the reconstruction processes and the decisions and justifications they contain leads to the loss of knowledge about rejected solutions. However, these contain valuable information that was developed during the reconstruction process. In the event of a change in the source situation or a re-evaluation of the sources used, the documentation of sources that were rejected or not considered as plausible would make it possible to start the reconstruction process anew and ideally make use of the earlier findings.

Actors have been aware of the problem of lack of documentation for a long time: in theoretical policy papers such as the London Charter [4] and Principles of Seville [5] (→ Scholarly Method) are formulated in terms of sustainability, verifiability, and knowledge preservation: “The documentation of evaluative, analytical, deductive, interpretative and creative decisions made in the course of computer-assisted visualization should be available in such a way that the relationship between research sources, tacit knowledge and explicit conclusions and visualization-based results can be understood” [6]. The working group on digital 3D construction for digital humanities in the German-speaking countries (DHd) sees the documentation as one of five core issues, as is reflected in their 2019 publication, Der Modelle Tugend 2.0 [7]. Despite this awareness, scientific documentation of digital reconstruction is an absolute exception even in 2024. Several reasons seem to converge here. To date, documentation is generally neither explicitly demanded by funding agencies, nor are additional funds made available for it. This applies to funding lines of the federal and state governments, but also to museum commissions, in that digital reconstructions often trigger or are created based on the latest state of research and with scientific advice. Thus, it is usually left up to the persons and institutions that create a reconstruction to finance documentation with their own funds. Furthermore, there is no agreement on documentation standards, structure, or content.

8.2.2 Preliminary Work

At the latest since the EPOCH Research Agenda [8] in 2004–2008, the topic of documenting 3D reconstructions has come into focus. One result of this four-year research project is the London Charter, which remains decisive for theoretical considerations and practical implementations. For some years now, there have been initial attempts in the professional community to meet the challenge of a lack of documentation with concrete solutions. These include a good overview of the ideas that have emerged up to 2017 [9] and a first draft for a tool for systematic documentation [10] with a web-based tool based on wiki technology (i.e., interlinked pages). The aim of the tool was to systematically document sources, their interpretation, hypothesis formation, and the resulting visualizations. However, the development never went beyond a prototype.

To better ensure data sharing, projects have focused on data modeling and linked data technologies in the context of digital 3D modeling—most notably the CIDOC Conceptual Reference Model (CRM) [11, 12], an ontology for cultural heritage, originally designed for use in museums. Within the framework of the project Virtual Reconstructions in Transnational Research Environments—The Portal: Castles and Parks in Former East Prussia [13], the first CIDOC CRM-referenced application ontology was developed to document the creative decision-making processes and the results of a digital (hypothetical) 3D reconstruction, including versions and variants. In follow-up projects at the Hochschule Mainz, the OntSciDoc3D application ontology has been successfully used in the semantic enrichment and labeling of digital 3D models with information on the provenance of the digital 3D dataset and the historical context of the represented object [14].¹

At the Hochschule für Technik und Wirtschaft Dresden, the online tool DokuVis was developed, which enables exchange between modelers and scientific advisors directly on the 3D model. Further features are the display of variants and versions, integration of project management, measurement of distances and surfaces, and the ability to generate section views of and interactively compare sources with the model [15].

Extensive theoretical considerations on the topic of documentation were already developed in 2010 at the Department of Digital Design at the Technische Universität Darmstadt (TUD), first preliminary work 2010 Mieke Pfarr [16]. This department started from the assessment that simple, intuitive, and user-friendly approaches are needed to move toward a practice of scientific documentation. In 2016, the Reconstruction Argumentation Method (RAM) was developed, which was transferred into an online-based documentation tool, ScieDoc,² in 2017 [17, 18]. The core of the RAM approach in ScieDoc is the division of a reconstructed building into different spatial areas. Each of these areas is represented by 2D images of the reconstruction (renderings), by images of the sources used, and by a textual argumentation explaining how the reconstruction was inferred from the sources. For each area, it is possible to map several variants. ScieDoc was used in 86 projects by 2022.

Currently, Building Information Modeling (BIM), the 3D modeling standard in the construction industry, is being discussed in the field of digital (hypothetical) reconstructions [19,20,21]. BIM enables to store project information and to assign customized properties to building elements. BIM supporting software ensures interoperability and reusability of the project data through the common data exchange format Industry Foundation Classes.³

With an archaeological approach, Demetrescu [22] sees virtual reconstruction as an extension of the findings in excavations and links them with virtual elements and sources. Prototypically, an interactive tool was developed for the visualization and exploration of the data in conjunction with the 3D models [9].

Many actors also dealt with partial aspects of documentation. Above all, the degree of uncertainty associated with incomplete sources and the wide scope for interpretation plays a prominent role. Various metrics were developed, such as the level of hypothesis [23], or one based on fuzzy logic [24], or a classification depending on the information content and the need for interpretation of sources. [25] The question of visualizing the different levels of uncertainty directly on the model has been discussed several times [26] (→ Shading Aspects).

Thus, although several approaches to documentation of decision-making processes exist, the proposed solutions should rather be seen as prototypes that cannot be used ad hoc by a broad community. In 2022, the DFG project IDOVIR started, which combines the approaches of the abovementioned projects ScieDoc and DokuVis and uses the RAM method developed at the TUD. It is free to use and takes 15 min to learn. As it is embedded in the system landscape of the University and State Library Darmstadt, the professional provision and continuation of the project results is guaranteed.

Beyond documentation of the decision-making processes, it is desirable to be able to permanently refer to a reconstruction and to cite it well in a scientific sense. For many reconstructions that are created for exhibitions or documentaries, permanent access to the results achieved is often not guaranteed. For this purpose, it would be useful to have freely accessible 3D repositories, which visualize 3D models and make 3D datasets available, document the corresponding metadata, and provide information on the often unclear legal situation regarding the (re-)use of results or models of a reconstruction. Complementary to the IDOVIR project in this context is the DFG-funded project DFG 3D-Viewer, which provides an infrastructure for publishing, locating, and displaying 3D models and records information on the circumstances surrounding both the digital reconstruction and the digital 3D model behind it.

Both DFG projects have agreed on common data structures and terms and are also seeking this understanding with other initiatives in the field of digital infrastructures. Both are briefly explained and recommended in the section on key projects at the end of this chapter.

One general challenge is that no new tool, method, or standard can solve the restrictions concerning the publication of the (historical) sources on which a reconstruction is based [27]. This would be necessary for scientific discourse, but the rights holders do not always allow this or demand fees, a circumstance that results in major restrictions and hurdles. General agreements would have to be reached here.

8.2.3 Current Developments

Documentation that enables computer-assisted work with the digitally available research data requires formalization of the knowledge in a human- and machine-readable form. This is primarily a matter of setting up rules for all data that clearly communicate the logic and meaning of that data to the computer, so that instead of data ruins, a homogeneous and consistent database is created.

As early as the 1980s, fundamental ideas and concepts for the description and documentation of art and architecture were developed. Foto Marburg provided an important basis for this with far-reaching considerations on the cataloging of cultural heritage [28], which led to their documentation and administration system MIDAS in 1989 [29].

Standards enable subject-specific classification of subject matter and make an essential contribution to the unambiguous indexing of cultural heritage. First and foremost, these are the controlled Getty Vocabularies, such as Art & Architecture Thesaurus (AAT), Getty Thesaurus of Geographic Names (TGN), Cultural Objects Name Authority (CONA), and Union List of Artist Names (ULAN), as well as ICONCLASS for iconographic indexing.

Further Reading: Classification Standards for Knowledge Modeling

• Controlled Vocabularies: A controlled vocabulary is an organized arrangement of words and phrases used to index content and/or to retrieve content through browsing or searching. It typically includes preferred and variant terms and has a defined scope or describes a specific domain [52, p. 12].

• Thesaurus: A thesaurus combines the characteristics of synonym ring lists and taxonomies, together with additional features. A thesaurus is a semantic network of unique concepts, including relationships between synonyms, broader and narrower (parent/child) contexts, and other related concepts. Thesauri may be monolingual or multilingual [52, p. 24].

• Taxonomies: A taxonomy is an orderly classification for a defined domain. It may also be known as a faceted vocabulary. It comprises controlled vocabulary terms (generally only preferred terms) organized into a hierarchical structure. Each term in a taxonomy is in one or more parent/child (broader/ narrower) relationships to other terms in the taxonomy. There can be different types of parent/child relationships, such as whole/part, genus/species, or instance relationships. However, in good practice, all children of a given parent share the same type of relationship. A taxonomy may differ from a thesaurus in that it generally has shallower hierarchies and a less complicated structure. For example, it often has no equivalent (synonyms or variant terms) or related terms (associative relationships) [52, p. 22].

• Ontologies: These are less commonly used than the above three standards for art information. In common usage in computer science, an ontology is a formal, machine-readable specification of a conceptual model in which concepts, properties, relationships, functions, constraints, and axioms are all explicitly defined [52, p. 24]. A main distinction is between application-overarching reference ontologies and the deriving application ontologies, which are purpose-specific applications of reference ontologies [13].

In cultural heritage, CIDOC CRM has prevailed as a reference ontology since the mid-1990s. This human- and machine-readable formalization of knowledge is based on around 100 classes (entities) and 150 relations (properties) that describe the essential facts and the context of an object, starting from an activity, e.g., the creation of a work of art (Fig. 8.1).

Fig. 8.1

(Image: AI MAINZ, 2019)

Graphical representation of the CIDOC CRM referenced application ontology OntSciDoc3D for mapping the knowledge of the 3D reconstruction process, https://www.ontscidoc3d.hs-mainz.de/ontology/, accessed on 1.2.2023

The demand for the worldwide networking of knowledge is being addressed by the development of the Internet into Web 3.0. The idea of a Web 3.0 network of knowledge was first presented in 2001 under the term Semantic Web by Tim Berners-Lee, the inventor of the World Wide Web [30]. It is based on a Resource Description Framework (RDF), a graph-based data format. The basic units of RDF are the triples consisting of subject, predicate and object. Data models such as CIDOC CRM [31] are used to describe individual basic units, which in turn enable the machines to read out the meaning of the digitally available data (semantics = meaning). If the RDF datasets (graph databases) are linked with other external, similarly structured online resources, a network of information is created that extends beyond the datasets themselves and is known as Linked (Open) Data.

The formalized and structured documentation of knowledge is the basic prerequisite for the processing information or making it available digitally. The proper handling of digital research data became an important concern of national education and research initiatives in the 2010s, which found expression in the endeavor to establish National Research Data Infrastructures (→Further reading: Infrastructure programs for 3D data). The new data culture formulated here [32] follows the general desire of academic circles for digital research data to be findable, accessible, interoperable, and reusable (FAIR) (→Legislation) [33].

Parallel to these developments in digital humanities, digital 3D modeling was examined in the engineering sciences. With the development of computer graphics since the 1960s and the arrival of powerful PCs in planning offices since the 1990s, efficiency increased because of improved interoperability and sustainability of digital 3D datasets. Regarding digital source-based 3D reconstruction, the approaches from architecture and spatial and urban planning represent interesting points of reference.

Since the mid-1990s, representatives from the construction industry, with software companies, have been developing a data exchange format that can store not only the geometry but also the object-based properties (property sets) and exchange them between different disciplines involved in construction. The resulting data exchange format, Industry Foundation Classes (IFC), form the common language of those involved in construction and introduce a new planning methodology that began in the 1970s and became known as Building Information Modeling (BIM) in the early 2000s [34]. The methodology is based on the communication of the parties involved by means of discipline-related 3D models (architectural design, construction engineering, MEP engineering, etc.), which are federated in a coordination model, as the Single Source of Truth. The IFC ensures a low-loss flow of information in the process and the information is mapped in the object-oriented 3D model, which has a different Level of Geometry, and Level of Information as well as accompanying documents (2D plans, photos, drawings, audio-/video-files, etc.) linked to it, depending on the required Level of Information Need. The increase in efficiency is thus based on an internationally recognized standard according to which projects are structured and BIM Execution Plans designed in response to Exchange Information Requirements [35].

In the early 2000s, a similar effort began in spatial and urban planning. As a result, an xml-based application schema for storing and exchanging digital 3D city models was introduced. City Geography Markup Language [36] enables digital 3D city models to be enriched with additional information that is of crucial importance for sustainable urban planning and administration of large scale build environment.

8.3 What Are We Dealing With

8.3.1 Model Types, Methods, and Data Formats

Digital 3D reconstruction deals with a wide variety of data that pose a particular challenge in terms of documentation. If we start from the 3D models, different types of models and modelling methods entail different properties [37]. The common subdivision introduces three types of models: wireframe, surface, and solid. The wireframe model is mainly used today for visualization purposes. Thus, the surface and solid models are the actual model types, which are essentially based on two different modeling methods. While the surface model consists of several independent surfaces that are not assigned to a concrete object, the solid model is a closed body that represents an object.

In terms of methods, object-based 3D modeling software solutions use Constructive Solid Geometry (CSG), which depicts completed solid models of the objects. This modeling methodology is preferred in industrial design, machinery, and construction (architecture and civil engineering, especially in BIM-supporting applications). The surface models that favor free-form modeling are used in animated films, product and architectural visualization, and urban planning. The associated modeling methodology is called Boundary Representation (B-Rep).

The manifold data formats for 3D datasets pose a further challenge. With regard to documentation, especially interoperability and thus the sustainability of data formats, a distinction is made between proprietary 3D model formats (e.g. C4D for Maxon Cinema 4D or 3DS for Autodesk 3D Studio Max) which are specific to software in which they were created, and data exchange formats (e.g., OBJ, DAE, STL, FBX, IFC, CityGML, X3D, and gITF), which are widely used and supported by many software applications [38]. The construction industry and urban planning use specific standards that integrate both the geometric information and the meaning of the objects or the surfaces that define them, such as IFC for BIM (ISO 16739–1:2018) or CityGML [39]. For web-based visualization of the 3D models, formats such as X3D and gITF were designed, which enable high-performance interactive display [40], but present challenges when converting the digital models from proprietary formats.

Finally, two trends should be briefly mentioned. First, standardization of methods and types of models, such as CSG/BIM and B-Rep/CityGML, can technically guarantee a documentation link and interoperability, are developing further and further, also regarding the graphic appearance. Second, increasing differentiation of the creation process for quasi-photorealistic reconstructions, e.g., immersive virtual reality simulations. A wide variety of software types are used, each of which fulfills specific functions to create geometries and represent textures, vegetation, or lighting conditions in generative models. However, the generative approach is not based on storing the resulting 3D geometries but on parameters and generates a 3D object in real time [41, 42]. In result storing 3D information in a state is not possible. With such modeling methods and creation processes, annotation of documentation directly on the 3D model is rather more difficult and storage in non-proprietary file formats may be impossible [43].

The question of a common language, the lowest common denominator of digital 3D models, may arise here. Agreement on a modeling, documentation, and publication method could guarantee the FAIR use of 3D models as serious 3D in archaeology, art and architectural history. Further considerations in this regard can be based on the concept of the Digital Critical Model [44], and the Scientific Reference Model [45, 51] (→ Scholarly Method).

8.3.2 Intellectual Argumentation

The documentation of decision-making processes is about recording trains of thought, interpretations of sources, and the considerations derived from them. Intellectual argumentation is supposed to explain—based on selected sources and analogies—why the reconstruction turned out the way it did, and which other possibilities were considered or discarded. If documentation exists at all, it is usually not in interoperable form. The situation is very inconsistent: apart from a few project descriptions online or in printed publications, documentation tends to exist internally, if at all. The spectrum ranges from handwritten records and/or the collection of sources used in file folders to embedding in complex internal communication tools that are based on commercial software and then developed by the users. Publishing/networking the data in this internal form is rather undesirable and seems technically difficult.⁴

In addition, some projects map decision-making processes with online databases and may link sources and considerations to 2D images of the reconstructions, or annotate them on the 3D models, and try to establish further references using Linked (Open) Data technologies. In almost all cases these are pilot projects that take up larger resources and have not yet been put to widespread use.

8.4 What Needs to be Documented

Documentation concerns the information about the created dataset and the decision-making processes involved in the reconstruction accompanied by object (physical and digital) related data. The information about a dataset is called metadata (data about data). From the long experience of recording objects, e.g., in a museum context, two main types exist: descriptive and administrative metadata. The essential data that is collected during documentation is known as the core dataset. Since 2006, the digital visualization of cultural heritage has included paradata [4]. By paradata, we mean the collection of information describing the process of creating the 3D model and the associated 3D visualization. First and foremost, it is the recording and documentation of the creative interpretation process that distinguishes the source-based 3D reconstructions from a retro-digitization of existing objects by terrestrial laser scanning or photogrammetry [46]. Here, the sources used, the gaps in knowledge uncovered, and the interpretation of the result are to be documented.

8.4.1 Documentation of the 3D Models: Metadata

The descriptive metadata describes the digital 3D model and the represented physical object. Usually, the name of the object, its function or type, its origin, its classification in an architectural epoch, the persons and events associated with it, etc. are mentioned.

The technical metadata related to the digital 3D model can include information on file size, file format, number of faces, edges, and nodes, model type, modeling methodology, software used, etc. This type of data can be partly mapped automatically during web-based 3D visualization.

The administrative metadata regulates the legal aspects of the object, in this case the digital 3D model. They name the author (creator) and possibly the persons involved, the persons and/or institutions as rights holders, and the license under which the 3D model can be used. They are the basic prerequisite for legally secure access to the 3D model and its future use.

As part of the development of infrastructure projects, core datasets for the documentation of 3D models, including digital-source-based 3D reconstruction, are currently being discussed and developed.⁵ The documentation schema presented here represents the comparison between IDOVIR and DFG 3D-Viewer and represents the status of the discussion from June 2022 (Fig. 8.2).

Fig. 8.2

(Image: AI MAINZ, 2022)

Miro board documentation scheme for alignment of core datasets between IDOVIR and DFG 3D Viewer

When capturing the digital 3D model, the data types can be classified according to the following pattern. The description of the dataset should contain the basic, mostly field-based information about the digital 3D model. Usually, this is the model’s name, the time period the model represents, and a free-text description of the model.

A central part of the documentation is the rights declaration, which clarifies the use of the model and should ideally be available under Creative Commons licenses. It contains field-based entries on the license, the author (creator), and other rights holders (e.g. legal bodies). It is important to refer to common international standards and norms regarding licenses (→ Legislation), persons, and entities (ORCID,⁶ GND,⁷ VIAF,⁸ etc.). In this way, ambiguities regarding machine information processing can be avoided and the datasets can be linked. To document the history of the model more comprehensively, the software used, the modeling method, other people involved, and time span of the 3D modeling can be recorded. Information on the physical (no longer existing) object is documented, which further contextualizes the model. Here, in addition to listing alternative names for the object, the object type can be defined following controlled vocabularies, and a link given to the Wikipedia and Wikidata datasets. Historical relationships to persons with a role and to historical events can ensure further contextualization of the object. The location of the object in relation to norm data, such as Geonames, should enable clear georeferencing of the object in question.

Finally, the project under which the 3D model was created can be documented within a project description. Here, it is useful to include the project title, possibly its acronyms, the website, the purpose, and the result as well as a free-text description of the project and the project duration including the institutions involved in the project and their respective roles (Fig. 8.3).

Fig. 8.3

(Image: AI MAINZ, 2022)

Documentation of the 3D reconstruction of the wooden synagogue in Volpa in the 3D repository developed in the DFG 3D-Viewer project.

8.4.2 Documentation of the Decision-Making Processes: Paradata

When documenting the decision-making processes, the intellectual argumentation behind them is central. It must be explained which sources, analogies, and considerations led to the reconstruction (Fig. 8.4). These explanations should be visually juxtaposed with the presented sub-area of a reconstruction in manageable spatial divisions (e.g., north façade, east façade, etc.) to enable immediate reference and comparison. The reconstruction can be visualized in a 2D image of the reconstruction model or by means of an annotation and a pre-set perspective on the 3D model. Both have different constraints and advantages. The 2D image does not require much effort to create. The representation of the 3D model may mean saving in a format suitable for the web, which may lead to a loss of representation or high effort but allows a good spatial clarification of the situation.

Furthermore, it makes sense to describe the variants considered and excluded in order to secure the knowledge stored in them and make it visible and, as described in the previous section, to include the metadata.

It could also be helpful to show the degree of uncertainty. Interesting work at the University of Bologna (→ Visualization) should be mentioned here [44], which is developing proposals on how the degree of uncertainty and sources used could be visualized on the 3D model or its representation.

Fig. 8.4

(Image: IDOVIR, 2022)

IDOVIR: documentation of the reconstruction of Altenberg Monastery

8.5 How to Document

Once it has been decided what is to be documented, there is the legitimate question of how to do so. This technical methodological aspect has so far been disregarded in the relevant guidelines and directives, not least due to the lack of research data infrastructures. As a result, 3D datasets could not be documented on a significant scale. New technological solutions were presented in the above section on preliminary work.

With funding from the German Research Foundation (DFG) and efforts from the German Research Data Infrastructures (NFDI), technological services are currently being developed and made available in the short term to enable the documentation and publication of research data, including 3D models, while adhering to the FAIR Principles and applying Linked Data technologies.

In analogy to Tim Berners Lee’s 5 Star Model for Open Data, the technical solutions for documenting 3D models within the 3D repositories are also described here in stages (Fig. 8.5).

Fig. 8.5

(Image: AI MAINZ/Igor Bajena, 2023)

Assessment schema for 3D formats in terms of their interoperability, based on the criteria of the 5-star deployment schema presented by Tim Berners-Lee in 2012

★ Provide your 3D model and the associated metadata on the web under an Open License (OL). The format does not matter (scanned sheet, PDF, etc.)

★★ Provide your 3D model in a format supporting Model Structure (MS) and the associated metadata on the web in a structured format (e.g. Excel instead of a scanned sheet, image or PDF)

★★★ Provide your 3D model in Neutral Format (NF) and use open, non-proprietary formats (e.g. CSV instead of Excel) for the associated metadata on the web

★★★★ Provide your 3D model with Structural Elements Properties (SEP) and use URIs to label things so that your data can be linked to

★★★★★ Provide your 3D model as Linked Open Model (LOM) and link your data with other data to create contexts (Linked Open Data).

This technical networked semantic documentation of 3D models (metadata) in 3D repositories is followed by documentation of the creative processes behind the visual representation (paradata). The striking feature of this context is the focus on recording the individual decision in a source-based reconstruction of art and architectural objects that no longer exist. Depending on the scientific questions of the research projects, what is to be documented may vary. Regarding standardization, jointly developed core datasets are of fundamental importance. The comprehensive documentation of decisions in a reconstruction process can also make use of the stage-like representation described above. The main difference here is that the 3D model does not necessarily have to be provided, as in the case of 3D repositories on the web.

Finally, it is important in the “how to document” that the hurdles are kept as low as possible e.g.: A cost-free infrastructure that can perhaps already be used as a communication platform during reconstruction. A minimal training time and as little additional effort as possible. Securing the data through renowned, experienced institutions such as libraries, can build trust. The core task remains to reveal the knowledge behind the reconstruction, whether in 2D or 3D. Ultimately, any form of documentation—publicly accessible or not—is better than no documentation at all.

Summary

After reading this chapter you should understand why it is important to document meta- and paradata of digital 3D reconstructions, what and how to document, and how to document projects with existing online tools.

Standards and guidelines

• UNESCO Charter (2003) Charter on the preservation of the digital heritage, published 2009 (https://unesdoc.unesco.org/ark:/48223/pf0000179529, accessed on 1.2.2023).

• DFG Guidelines for Safeguarding Good Scientific Practice (2019), https://www.dfg.de/download/pdf/foerderung/rechtliche_rahmenbedingungen/gute_wissenschaftliche_praxis/kodex_gwp.pdf, accessed on 1.2.2023; https://www.dfg.de/download/pdf/foerderung/rechtliche_rahmenbedingungen/gute_wissenschaftliche_praxis/kodex_gwp_en.pdf, accessed on 1.2.2023.

• DFG Code of Practice “Digitization” (2016) https://www.dfg.de/formulare/12_151/, accessed on 1.2.2023.

• London Charter, Denard H (2012) A new introduction to the London charter. In: Bentkowska-Kafel A et al (eds) Paradata and transparency in virtual heritage. Ashgate, Burlington, pp 57–72. https://www.londoncharter.org/fileadmin/templates/main/docs/london_charter_2_1_de.pdf, accessed on 1.2.2023.

• Principles of Seville. International Principles of Virtual Archaeology. Ratified by the 19th ICOMOS General Assembly in New Delhi, December 2017 (http://sevilleprinciples.com, accessed on 1.2.2023).

• FAIR Principles, Wilkinson et al (2016) The FAIR guiding principles for scientific data management and stewardship. Scientific Data 3(1):160,018. https://doi.org/10.1038/SDATA.2016.18.

Projects

• DFG 3D-Viewer (2021–2023). The development of a 3D viewer infrastructure for historical 3D reconstructions is intended to provide permanent publishing and archiving of 3D datasets and the associated metadata, and to enable collaboration and expert discourse on digital 3D models. The overall goal of the project is to develop a web-based 3D viewer and exchange format for distributed 3D repositories that ensure findable, accessible, interoperable and reusable 3D models as research data. Subgoals include creating an interdisciplinary application profile, including the assignment of rights (licensing of models) and developing a prototype 3D repository and workflow for delivering a 3D model into 3D repositories. https://dfg-viewer.de/en/dfg-3d-viewer, accessed on 1.2.2023.

• DFG Project IDOVIR: Infrastructure for the documentation of virtual reconstructions (2022–2024). The aim of IDOVIR is to make decision-making processes in architectural reconstruction permanently traceable and openly accessible online with embedding at the University and State Library Darmstadt. The core is the presentation of the intellectual argumentation behind a reconstruction. Special attention is paid to the ability to present any number of variants or working states. Other features of the research infrastructure, which can be used free of charge, include low-threshold use (familiarization time 15 min), guided data entry, automatic PDF generation of the documentation, visualization via 2D renderings and an interactive 3D model, and various tools for evaluation, measurement, and communication. Depending on the available resources, it allows minimal or extensive documentation. IDOVIR also developed a proposal on how to depict the degree of uncertainty (plausibility) of the reconstruction and the sources used. https://idovir.com/, accessed on 1.2.2023.

• Semantic Kompakkt: A tool for the semantic annotation of 3D models, developed within NFDI4Culture (2020–2025). The open-source components Wikibase (indexing, knowledge graph), Open Refine (data transformation and import) and Kompakkt (storage, visualization, interaction) will be integrated. Initially, for the use case of cultural heritage and its specific vocabularies and ontologies such as CIDOC CRM, this enables the annotation of 3D data within an open and flexible knowledge graph environment. This facilitates the linking of 3D models, annotations, and their cultural context with the semantic web and with national and international standards data. Semantic Kompakkt is designed as a collaborative environment with different levels of read/write access, where research groups can use graphical user interfaces to upload and annotate data with a clear provenance and metadata.

  https://gitlab.com/nfdi4culture/ta1-data-enrichment/kompakkt-docker, accessed on 1.2.2023.

Key literature

• Grellert, M., Pfarr-Harfst, M., 2019. The Reconstruction Argument Method—Minimum Documentation Standard in the Context of Digital Reconstruction. In: Kuroczyński, P., Pfarr-Harfst, M., Münster, S. (eds.), Der Modelle Tugend 2.0: Digitale 3D-Rekonstruktion als virtueller Raum der architekturhistorischen Forschung, arthistoricum.net, Heidelberg, pp. 264–280 [47].

• Wacker, M., Bruschke, J., 2019. Documentation of digital reconstruction projects. In: Kuroczyński, P., Pfarr-Harfst, M., Münster, S. (eds.), Der Modelle Tugend 2.0: Digitale 3D-Rekonstruktion als virtueller Raum der architekturhistorischen Forschung, arthistoricum.net, Heidelberg, pp. 282–294 [48].

• Apollonio, F.I., 2016. Classification Schemes for Visualization of Uncertainty in Digital Hypothetical Reconstruction. In: Münster, S., Pfarr-Harfst, M., Kuroczyński, P., Ioannides, M. (eds.), 3D Research Challenges in Cultural Heritage II: How to Manage Data and Knowledge Related to Interpretative Digital 3D Reconstructions of Cultural Heritage, Springer International Publishing, Cham, pp. 173–197 [25].

• Kuroczyński P., Bajena I.P., Große P., Jara K., Wnęk K., 2021. Digital Reconstruction of the New Synagogue in Breslau: New Approaches to Object-Oriented Research. In: Niebling F., Münster S., Messemer H. (eds.): Research and Education in Urban History in the Age of Digital Libraries. UHDL 2019. Communications in Computer and Information Science, vol 1501. Springer, Cham, pp. 25–45 [49].

• Apollonio F.I., Fallavollita F., Foschi R., 2021. The Critical Digital Model for the Study of Unbuilt Architecture, in: F. Niebling et al. (Eds.): UHDL 2019, CCIS 1501, pp. 3–24 [50].

• Kuroczyński P., Apollonio F.I., Bajena I., Cazzaro I., 2023. Scientific Reference Model – Defining standards, methodology and implementation of serious 3D models in archaeology, art and architecture history, in: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLVIII-M‑2 – 2023, pp. 895–902 [51].

• Bajena I.P., Kuroczyński P., 2023. Development of the Methodology and Infrastructure for digital 3D Reconstructions, in: Proceedings of (IN)TANGIBLE HERITAGE(S) A conference on technology, culture and design, Canterbury 2022, AMPS conference proceedings series, ISSN 2398–9467, pp. 72–83 [45].