What are reusable data?

Efforts to make datasets accessible and reusable aim to:

  • make available to teams, customers, providers, citizens, etc. with the key datasets of an organization,
  • facilitate the reuse of datasets by applications, algorithms, statistical analysts …

The principles of accessibility and reusability are laid out in detail in the guidelines for “FAIR Data” (Findable, Accessible, Interoperable, Reusable).

The principles of accessibility and reusability apply as much to data open to the general public as they do to data with restricted access.

Accessibility

Information sheet for a dataset and a catalog of reusable datasets
A dataset description notice and a catalogue of reusable datasets

Ease of access to datasets relies on the standardized description of each dataset and the implementation of a catalog of datasets accessible via intranet or the web

Reusability

reusable and linked datasets
Two reusable, linked datasets

A dataset’s ease of reuse depends on various criteria:

  • providing users with the organizational model of the data and the documentation of the dataset,
  • providing the data in readable, standardized formats (JSON, CSV, XML, RDF …),
  • working on the interoperability of the data to facilitate their aggregation: 
    • when a value depends on a controlled vocabulary (reference concept), the latter must also be accessible and reusable,
    • when a value makes reference to a resource described in another dataset, the link must be explicit and the resource identified distinctively.

Use of Semantic Web Standards

The application of Semantic Web standards facilitates the implementation of the principles of dataset accessibility, interoperability and reusability thanks to the identification of resources by URIs, the publication of controlled vocabularies in accordance with the SKOS model, and the use of the DCAT model for describing datasets