ISIDORE is a search platform allowing access to digital data of Humanities and Social Sciences. Open to all, it relies on the principles of Web of data and provides free access to data (open access). More than a simple search interface, ISIDORE standardizes and enriches the metadata and data collected thanks to recognized vocabularies in three languages (French, English and Spanish).
The TGIR also launched a partnership with the Canadian Cyber Infrastructure on the "ISIDORE on Demand" service and organised a hackathon in Montreal in 2017, making collaborations even more concrete with French speaking research communities.
The contribution enables the members of DARIAH to promote their data on the ISIDORE platform by offering, according to the connectors of the latter (OAI-PMH and sitemap+RDFa), data set in open access. The eligible resources may be in French, in English and in Spanish. The enrichments will be in French, in English and in Spanish. The ISIDORE interface will also be in these three languages. The describing vocabularies of the metadata are: Dublin Core Element Set, DCTerms.
ISIDORE uses referentials to enrich the data: the MORESS referential, the index of the thematic categories of HAL and Open Edition platforms, the Rameau vocabulary, the Pactols, GEMET, GeoEthno thesauri and finally the Geonames geographical referential. Several enrichments are made in “categorising” the data for linking data to a scientific discipline or issue and an adjunction of terms derived from the diverse referentials. To conduct these enrichments, ISIDORE exploits the metadata of the resources as well as the integral text, by analysing these data to link them with the referentials (in the three languages). The elements present in the metadata or the integral are compared with the entries of the referentials through an algorithm based on a morphological analysis of the terms. If there is an equivalence between a term derived from the resource and an entry of one of the referentials, then the resource will be linked to said entry of the referential.
A chain of harvesting, enrichment and transformation into RDF of the describing metadata is made available in the context of the DARIAH project. In this context, the free access documents will be indexed as integral texts. The harvested data are accessible via the multilingual web ISIDORE interface, via the API and RDF endpoint SPARQL of ISIDORE. ISIDORE is a service provided by Huma-Num.