CAA2014 L'ontologie du CIDOC CRM pour interroger une base de données d'Archéologie

Interoperability of the ArSol (Archives du Sol) database

with the CIDOC-CRM ontology

E. Le Goff, O. MarletX. Rodier, S. Curet, Ph. Husi

Laboratoire Archéologie et TerritoiresUMR7324 CITERES

With the collaboration of: P. Le BœufBibliothèque Nationale de France

CAA2014PARIS

01/15

• Objectives of ArSol: data management, processing and analysis

▪ ArSol = Archives du Sol (DBMS : 4e Dimension)

▪ data management and research

• Background to the interoperability of digitized resources

▪ context of the MASA Consortium

(MASA = Mémoire des Archéologues et des Sites Archéologiques)

▪ funded by the TGIR HumaNum (Humanités Numériques)

▪ international standard CIDOC-CRM (ISO 21127:2006)

• Implications for ArSol and committed procedures

▪ bringing ArSol on line

▪ mapping: matches between CIDOC-CRM classes and properties,

and database fields in ArSol

CAA2014PARIS

02/15

INTR

OD

UC

TION

CIDOC-CRM ontology: between abstraction and reality

▪ Spatio-temporal definition of archaeological entities

- archaeological entities → recording → research archive

- interpretation with features and artefacts

▪ The need to appropriate the CIDOC Conceptual Reference Model

I. UN

DER

STAN

DIN

G O

NTO

LOG

Y

CAA2014PARIS

03/15

Mapping the FAIT and MOBILIER tables

in the ArSol database

▪ Selection of basic archaeological information

- information common to all

excavations: features and artefacts

- retain only the data characterizing the

feature or object

CAA2014PARIS

04/15

II. STEPS OF M

APPIN

G

▪ Mapping the ArSol fields on the CIDOC-CRM entities → correspondance

CAA2014PARIS

05/15

II. STEPS OF M

APPIN

G

▪ Mapping the ArSol fields

on the CIDOC-CRM

entities → RDF triples

CAA2014PARIS

06/15

II. STEPS OF M

APPIN

G

▪ Abstract model of data: FAIT’s particularities

CAA2014PARIS

07/15

II. STEPS OF M

APPIN

G

▪ Abstract model of data: FAIT’s particularities

CAA2014PARIS

07/15

II. STEPS OF M

APPIN

G

▪ Abstract model of data: MOBILIER’s particularities

CAA2014PARIS

09/15

II. STEPS OF M

APPIN

G

▪ Abstract model of data: MOBILIER’s particularities

CAA2014PARIS

09/15

II. STEPS OF M

APPIN

G

▪ Problems encountered

- archive vs resources:

divergent concepts

- the semantics specific

to archaeology

- how control the

conformity of the

mapping?

CAA2014PARIS

11/15

II. STEPS OF M

APPIN

G

III. INTER

OPER

AB

ILITYMeeting the needs of interoperability

• A script to check the

efficiency of the mapping

CAA2014PARIS

12/15

DBMS(ArSol server)

ReasonerProtégé-ontop-

Query applicationSPARQL Endpoint

MappingOBDA model

OntologyCIDOC-CRM(OWL/RDF)

SparQL

Query rewriting

Local client

Remote client(VPN)

Web client(Browser)

Web server(ArSol web)

Source

• Bring ArSol to the semantic web

- Web access

- External interrogation with OBDA system

- Database kept in its current form

Web client(Web service)

CAA2014PARIS

13/15

III. INTER

OPER

AB

ILITY

CAA2014PARIS

14/15

• Accessibility

of the

database

III. INTER

OPER

AB

ILITY

• Specific mapping for the ArSol database

• Methodology can be shared

• Semantic homogeneization

- CIDOC-CRM to structure information

- PACTOLS to standardize vocabulary

• Web-semantic framework

→ Toward a RDF version with the SKOS format

Acknowledgments

▪ P. Le Bœuf (BNF, Paris), for his invaluable and tireless help in the mapping phase.▪ A.-V. Szabados (CNRS, ArScan, Paris), R. Letricot (LARHRA, Lyon) and A. d’Andrea (Centro Interdipartimentale di Servizi di Archeo- logia, Italy) for sharing their experience of using the CIDOC CRM.▪ B. Bouchou Markhoff (Université Fr. Rabelais, Laboratoire d’Informatique, Tours) for recommending the use of ODBA and -ontop-, and for her commitment to the development of this project.

15/15Émeline Le Goff, Olivier Marlet – Laboratoire Archéologie et Territoires, UMR 7324CITERES / Tours

CAA2014PARIS

CO

NC

LUSIO

N