The potential of data to be retrieved and accessed. To be accessible according to the FAIR principles that data or metadata must be retrievable through a standardised protocol by resolving a globally unique identifier.

Controlled Vocabulary

A vocabulary which only includes terms that have been deliberately selected by its creators or maintainers.

Data Management Plan (DMP)

A guide describing a series of steps needed to handle data to meet one or more aims. For research data, a data management plan specifies what measures will be taken throughout and often beyond a research project to ensure that its data is well documented, findable, accessible, and reusable.

Digital repository

A place where data, code, software, or other digital assets are stored.


The property of data, if they are findable, accessible, interoperable and re-usable according to the FAIR principles

FAIR digital object

A digitial container for data or links to data which has its own identifier (usually unique, persistent and web accessible) and contains metadata describing the data and/or the FAIR digital object itself. The container itself is findable, accessible, interoperable, and reusable according to an implementation of the FAIR principles.

FAIR principles

A set of requirements that data properties needs to fulfill in order for the data to be findable, accessible, interoperable and re-usable. Thereby the data will be re-usable for humans and machine-agents in the long-term and thus be sustainable.


Data should be easy to find both by humans and by computer systems. (Source).

Requirements are meaningful metadata, IDs and open access (not necessarily part of the definition).


The data should be able to be exchanged, interpreted and at least semi-automatically combined with other data sets (Source).

The prerequisite for this is a common format or a kind of crosscompiler between different formats (this is not part of the definition, but it seems to be important).


Any data which is intended to help contextualise or otherwise qualify other data. Examples include file size, date created, method used to generate the data and identifiers for people who worked on the data.


A structured collection of terms, their definitions and the relationships between them expressed in a formal, typically machine readable, manner.

Persistent Identifier

A persistent identifier (PID) is a long-lasting reference to an object. Some persistent identifiers are linked to kernel information profiles which provide metadata or information about that object when the persistent identifier is resolved.


The history which is the sum of those processes relating to the origin, ownership, custody, location, and modification of an entity.


The data is reusable for future research and comparable with other data sources. Citability and terms of use also fall under this keyword (Source).

This implies, among other things, intensive use and provision of metadata such as test environment, version documentation, etc.