FAIR principles

What is FAIR Data?

One of the major challenges in storing research data is the optimal processing for humans and machines. The FAIR Data principles are intended to help with this.

FAIR stands for: Findable- Accessible- Interoperable- Re-usable

Findable: Data should be easy to find for humans and computers alike. This is achieved, among other things, by

• a unique and persistent identifier (e.g. DOI)
• detailed documentation of the context of the data (origin, further use, etc.) through metadata
• if possible use of standardized metadata and controlled vocabularies

Accessible: Data should be accessible in the long term and retrievable by humans and machines. This is achieved, among other things, through

• Access via standardized, open, free protocols
• clearly defined post-use conditions (not automatically open/free!)
• appropriate authentication and rights management functionality, if necessary
• long-term storage of metadata

Interoperable: Data should be technically reusable and - by man and machine - combinable with other data sets. This is achieved, among other things, by

• use of a formal language that can be interpreted by humans and computer systems alike
• interoperable and documented controlled vocabularies, thesauri and ontologies

Re-usable: Data should be analytically/intellectually re-usable, i.e. understandable and interpretable. This is achieved, among other things, by

• descriptive documents for intellectual classification of the data
• clear description of the terms of use
• provenance information as the basis for an evaluation of the data
• preparation according to (professional) standards

<<back