A framework stating that data should be Findable, Accessible, Interoperable, and Reusable — to maximize its value for both humans and machines.
The FAIR principles were published in 2016 and have become a widely adopted standard for research data management and open data publishing. Each principle addresses a different barrier to data reuse:
- Findable: Data has a persistent identifier and rich metadata so it can be discovered.
- Accessible: Data can be retrieved using open, standard protocols, with clear conditions for access.
- Interoperable: Data uses standard formats and vocabularies so it can be combined with other datasets.
- Reusable: Data has a clear licence and sufficient documentation to be used by others for new purposes.
Example: A federal research dataset is FAIR if it has a DOI (findable), is downloadable via HTTPS (accessible), uses standard column names and codes (interoperable), and is published under an open government licence with a data dictionary (reusable).
FAIR does not require that data be open — some FAIR data is restricted to authorized users. But open data should always aim to be FAIR.