Display, Ontology and Database for Exhibition Documentation

Renaudie Zoë, Château-Dutier Emmanuel, Krause Lena, Valentine David[¹]

Introduction

The Display project, incubated within L’Ouvroir, the Digital Art History and Museology Laboratory of the Université de Montréal (Canada), emerges as a collaborative endeavor under New Uses of Collections in Art Museums Partnership². To document the history of hanging collections in art museums and be able to reconstruct them, the software solution that we are developing must make it possible to support all research operations, from the collection of historical information, formulating hypotheses and recording results. This initiative, driven by a multidisciplinary team composed of art historians, computer scientists, and museologists, aims to develop a digital tool tailored to streamline research on exhibition displays. In this poster we will present our methodology in shaping this tool’s conception and underscore the pivotal role of the DH Lab in the project.
In what follows, we first discuss who we are and the context of the creation of this initiative, then what Display is meant to be, and to finish we present our methodology and results.

Research Context

The evolving landscape of museum strategies has witnessed a paradigm shift, marked by an influx of innovative initiatives permeating collection curation. From anachronistic intrusions to thematic presentations, these strategies underscore a departure from traditional exhibition norms. In this context was created, CIÉCO to bring together the main axes of study in francophone museology in Quebec and in Canada. Its New Uses of Collections in Art Museums Partnership (2021-2028) brings together the museum and the university in a co-production of knowledge on the transformation of practices surrounding museum collections. The partnership’s collaborative ethos is manifested through annual research workshops, fostering knowledge co-construction among principal investigators, museum professionals, and young researchers. L’Ouvroir was partly created to provide this partnership with digital tools. Member of the Interuniversity Research Center on Digital Humanities, CRIHN, this DH lab provides Canada with a dedicated research facility for experimentation and the development of innovations in the fields of digital art history and museology. In this context, the Ouvroir team was given three years to develop three tools, one of them being Display.

Project Concept

Central to the CIÉCO project is the mobilization and utilization of extensive archival sources to meticulously document the evolution of art museum exhibitions and facilitate their reconstruction. The essence of the Display project lies in crafting a tool that aids in all research facets, from gathering historical data to formulating hypotheses and documenting findings.
The evolving landscape of digital tools for art historians underscores the need for streamlined methodologies and intuitive interfaces. Digital conservation of exhibitions is mainly 3D rendering of the exhibition spaces or online diffusion of archival documents (exhibition views, list of artworks, plans…) (Schweibenz et Scopigno 2018). Our tool is intended to support the research team in their archive consultation for documenting exhibition displays. It should enable them to maximize the complex modeling of an exhibition based on potentially incomplete information. The point of our project is to offer art historians an easy way to visualize the spatial arrangement of exhibits diagrammatically or in 3D, even with incomplete information.
One of the advantages of the software is that it allows collective work to be organized on the documentation of exhibitions. The aim is to provide a tool for research assistants and researchers to work on exhibitions. The application is intended to record the information collected, in particular in exhibitions archives. We imagined three principal types of users:

• Research Assistant: The database aids research assistants by organizing their work for principal investigators. They sift through various exhibition archives, extracting location information on exhibited objects. Documentation includes lists of works, exhibition views, plans, texts, press clippings, and invoices. Handling abundant information, assistants discern correct data and translate it spatially. They input information sequentially, using the application to identify source-related issues for the researcher. Visualization clarifies exhibition understanding and helps identify conflicting information.
• Principal Investigator: Researchers utilize the application to formulate restitution hypotheses based on incomplete information or hypotheses from assistants. They test and confront solutions, employing a heuristic approach to reason about available information and seek feedback. Visualization aids in proposing restitution hypotheses, allowing researchers to isolate spaces for focused modeling, such as around a room or specific artwork across multiple projects.
• Curator: Curators may use the application to design exhibitions, supporting the creation process before production. Visualization aids in proposing scenarios, and curators rely on visual documentation to discuss projects and reference other works easily. They aim to produce preliminary documents for exhibition setup or for scenographer work, grouping works by subject or room without strict arrangement precision.

The members involved in producing the project are from multiple grounds and have different roles. The project had three principal steps:

1. Determine what exists//what is needed: discussions and workshops between Users Team and DH Team
2. Conceive the tool: DH Team
3. Produce the tool: external developer and DH Team

After the first step, a global vision of what was needed emerged: a digital tool but also a model of thought. The initial phase of conceiving the tool involved designing a computer ontology (Gruber 1993), meticulously detailing the characteristics of an exhibition display. Some important domain ontologies exist like CIDOC-CRM (Bekiari et al. 2022), and we took a certain interest in its applications (Carboni, Usel, et Joyeux-Prunel 2023), or in Linked Art (Delmas-Glass et Sanderson 2020) and in OntoExhibit (del Mar Roldán García et al. 2022), but none express what we want to represent, the way we want to represent it. To incorporate topography, we studied a few ontologies in domains like geography, archeology, geometry but the one that was the most interesting for us was from architecture: Building Topology Ontology (BOT) (Rasmussen et al. 2021). Other than ontologies we studied RCC8 involving an architectural perspective using CIDOC, CRMgeo, and GeoSPARQL for describing the topological relationships between elements of a building. The principles underlying the choices and tactics are very similar to those of Display (Guillem et al. 2023).
We had intense meetings, between members of the lab, learning from each other, to imagine together how to represent this model and what vocabulary to use. L’Ouvroir’s members are both in art history and in digital humanities, part of our role is to translate to either the art historians or the technical developers what each needs to understand. The ontology is one of our ways to do so.

Ontology for displays

Figure 1: Core classes of the Display Ontology (figure by the authors)

In its current state, our computer ontology is intended to describe in an explicit and formal way the features of a collection display or an exhibition (identification of the exhibition, proximity and contiguity of the works, vis-à-vis, etc.). It is aimed to facilitate the recording of historical information about hangings and proposes solutions to account for uncertainty and documentary gaps. We want it to be compatible with the CIDOC-CRM ontology to cover the specific domain of displays and exhibitions. If we didn’t involve outside stakeholders yet, we want the resulting model to be published for comments (Request for Comments) within the museum documentation community and we consider an iterative process to make it evolve. All our work and projects are accessible on GitHub ouvroir: https://github.com/ouvroir.
The functionalities offered by the data model and the application are based on a spatial approach and the use of 3D modeling to facilitate the reconstruction of collection displays. The system exploits the data to propose calculated reconstructions based on the analysis of the spatial constraints of the exhibition spaces and information on the works exhibited (dimensions, weight, lighting). Given the often scant documentary landscape, our abstract model adopts a spatial approach to capture historical insights into exhibition installations. This model enables the recording of crucial information by delineating topological inferences between objects within the exhibition space. We tested the ontology on an exhibition in order to take to modelling process further. The exhibition being Feux Pâles (production of readymades belong to everyone® and CAPC, 1992), we had access to archives as I did the archival work on (Renaudie 2020). This raised some interesting questions we had to answer. The next step is to test it with other examples from the CIECO partners.

The Software

Furthermore, deliberations regarding the database tool form the project’s secondary research axis. The integration of the database and user interfaces must be carried out by an external developer. Display entails not only the creation of a robust backend infrastructure but also a user-friendly frontend interface tailored to accommodate diverse user profiles. Through the synergistic interplay between backend and frontend components, we aim to furnish art historians with a versatile toolset conducive to their research. With the utilised document model, this information should be qualified by level of evidence and sourced, and also shared as linked open data. One of our main focuses is inference management to help the researcher. We are in this phase of discussion around the organisation of a RDF back-end with a JSON front-end. However, such a solution is not easy to implement, the fields of UX design and semantic web technologies being relatively disjointed. We want the triple-store for the inferences and an UX web-based interface to facilitate tasks such as artwork list creation, exhibition space geometry definition, and artwork localization. Therefore we are currently considering the thickness of the abstraction layer the interface should have.

Conclusion

In conclusion, the Display project and its collaborative efforts show how we work between different domains and different interests. The DH lab in this case is the mediator between the humanities and digital. We welcome people interested in the project to reach out at ouvroir@umontreal.ca.

Acknowledgments

This work is part of the research of the New Uses of Collections in Art Museums partnership funded by the Social Sciences and Humanities Research Council of Canada : https://cieco.co/fr

References

Bekiari, Chryssoula, George Bruseker, Erin Canning, Martin Doerr, Philippe Michon, Christian-Emil Ore, Stephen Stead, et Athanasios Velios. 2022. « Volume A: Definition of the CIDOC Conceptual Reference Model ».
Carboni, Nicola, Thibault Usel, et Béatrice Joyeux-Prunel. 2023. « Intégrer des données historiques spatio-temporelles ». In . https://hal.science/hal-04110759.
Delmas-Glass, Emmanuelle, et Robert Sanderson. 2020. « Fostering a Community of PHAROS Scholars through the Adoption of Open Standards ». Art Libraries Journal 45 (1): 19‑23. https://doi.org/10.1017/alj.2019.32.
Gruber, Thomas R. 1993. « A translation approach to portable ontology specifications ». Knowledge Acquisition 5 (2): 199‑220. https://doi.org/10.1006/knac.1993.1008.
Guillem, Anaïs, Antoine Gros, Kevin Reby, Violette Abergel, et Livio DeLuca. 2023. « RCC8 for CIDOC CRM: Semantic Modeling of Mereological and Topological Spatial Relations in Notre-Dame de Paris ». In SWODCH 2023, édité par Antonis Bikakis, Roberta Ferrario, Stéphane Jean, Béatrice Markhoff, Alessandro Mosca, et Marianna Nicolosi Asmundo. Athens, Greece, November 7, 2023. https://ceur-ws.org/Vol-3540/paper2.pdf.
Mar Roldán García, Maria del, Martin Jerónimo Salvachúa, María Luisa Díez Platas, et Nuria Rodriguez Ortega. 2022. « OntoExhibit ». Ontology Specification Draft. 2022. https://w3id.org/OntoExhibit/.
Rasmussen, Mads Holten, Pieter Pauwels, Maxime Lefrançois, et Georg Ferdinand Schneider. 2021. « Building Topology Ontology ». Draft Community Group Report. W3C. https://w3c-lbd-cg.github.io/bot/.
Renaudie, Zoë. 2020. « The world of Feux pâles: An exhibition put to the test of conservation ». ICAR, no 4: 205‑17.
Schweibenz, Werner, et Roberto Scopigno. 2018. « Documenting Past Exhibitions: Why and How Information Technology Could Help to Preserve Dismantled Shows ». Uncommon Culture 7 (1/2, Cultural Heritage, Real\&Virtual): 75‑85.