On using the CIDOC CRM to model archaeological datasets - HAL Accéder directement au contenu
Communication dans un congrès Année : 2021

On using the CIDOC CRM to model archaeological datasets

Utiliser l'ontologie du CIDOC CRM pour modéliser des jeux de données archéologiques

Résumé

The rise of the Semantic Web and the opening of research data in the Open Science dynamic are encouraging French archaeologists to apply FAIR principles (Findable, Accessible, Interoperable, Reusable data) to make their data sustainable, interoperable and reusable. This is one of the main missions of the MASA consortium (Memories of Archaeologists and Archaeological Sites) of the TGIR Huma-Num: to support archaeologists in this process by relying on a standard ontology massively shared by a large part of the Cultural Heritage actors at the international level, the CIDOC CRM. This paper’s goal is to present how modeling with the CIDOC CRM can be adjusted according to the corpus. The richness of this ontology allows indeed to set up models at different granularities in order to reach the according objectives. We propose to compare two experiences made in two research projects: the first one relies on a generic model for several archaeological datasets within the MASA Consortium; the second one requires a more specific model for excavation data and architectural data within the framework of the SESAMES project (Semantization and Spatialization of Multi-Scale Heritage Artifacts: 3D annotation, Sonification and formalization of reasoning). It is important to notice that both of these models are extracted from the CIDOC CRM and its extensions, without defining any new concept or property. This is a choice that we believe is required in order to be consistent with the best practices of the semantic Web and the FAIR principles. A generic model for the whole archaeological corpus. Bringing the different archaeological datasets of the MASA consortium to the semantic web required to set up a triplestore structured with a common model. This generic model, based on the CIDOC CRM, provides a unique and transversal query interface for heterogeneous archaeological datasets. These ones are published on the semantic Web via the OpenArchaeo platform (Marlet et al. 2019). More precisely, the OpenArchaeo generic model is based on the CIDOC-CRM ontology and some of its extensions (CRMarchaeo, CRMsci and CRMba). This model relies on basic concepts found in most archaeological datasets (site, excavation, structure, feature, wall, burial, stratigraphic unit and artifact). While this model was designed to be as generic as possible, we find that some datasets cannot use all its concepts and relationships. In fact, only one concept is necessarily found in all archaeological datasets, that of the site. This concept, as defined in the CIDOC CRM (http://www.cidoc-crm.org/Entity/e27-site/version-6.2.1), is the common dominator in all archeological datasets. The development of OpenArchaeo generic model is based on the work of the MASA consortium within the European program ARIADNEplus. The MASA consortium aims to enrich the ARIADNEplus platform with data integrated in OpenArchaeo. A detailed model for archaeological fieldwork The SESAMES project seeks to experiment and to share several semantic approaches of multi-scale built features (from the archaeological fragment to the building) in three of their dimensions (spatial, sound, ontological). By using formal ontologies, the SESAMES program focuses on the articulation of two different views of the same heritage items: that of architects and that of archaeologists. The CRMarchaeo extension of the CIDOC CRM, the ARIADNEPlus Application Profiles and the OpenArchaeo generic model testify of the ability of CIDOC CRM to represent the archaeological approach. Moreover, its CRMba extension breaks down built structures into both functional and construction units, which can be phased according to the principles of stratigraphy (Ronzino, Niccolucci, Felicetti et al. 2016). The CIDOC is thus adapted to make the items observed by archaeologists and those observed by architects consistent with each other, while these items may be defined at different spatial and chronological scales. While the OpenArchaeo model is generic enough to connect archaeological datasets on various scales, the model developped for SESAMES is much more detailed. Its specificity should bring together the architectural study with the building archaeology. The produced chronological interpretation shall be explicited. Moreover, beyond the design of a model fitted for built heritage, this work enables us to evaluate the relevance of the implementation of a model based on the CIDOC CRM applied at a finer scale. It is therefore completely in line with the reflection on the Application Profiles of ARIADNEplus, in particular the "Excavation Archives" one (Katsianis, Kotsakis and Stefanou 2021), whose objective is to represent the way in which data is enriched and reorganized into distinct datasets during their successive interrogations.
Fichier principal
Vignette du fichier
CAA2021 - On using CIDOC to model archeological datasets.pdf ( 1.26 Mo ) Télécharger
Origine : Fichiers produits par l'(les) auteur(s)
Licence : Paternité - CC BY 4.0
Loading...

Dates et versions

halshs-04199967, version 1 (08-09-2023)

Licence

Paternité - CC BY 4.0

Identifiants

  • HAL Id : halshs-04199967 , version 1

Citer

Olivier Marlet, Théo Roulet, Florian Hivert, Béatrice Markhoff, Xavier Rodier, et al.. On using the CIDOC CRM to model archaeological datasets. CAA2021 - Digital Crossroads, Jun 2021, Limasolle (virtual), Cyprus. ⟨halshs-04199967⟩
36 Consultations
50 Téléchargements
Dernière date de mise à jour le 27/04/2024
comment ces indicateurs sont-ils produits

Partager

Gmail Facebook Twitter LinkedIn Plus