Skip to content
Computer programmer coding on modern computer with AI Artificial

Publication & archiving

Once collected, selected and processed according to the standards set out in the DMP (= data management plan) during the planning phase, the data must be published and archived for the long term.* Compliance with the FAIR principles should be checked in detail at the very latest at this stage. All data considered relevant and intended for long-term archiving should be Findable, Accessible, Interoperable and Reusable (FAIR):

Findable

Findability is ensured by enriching the data with metadata, in particular by assigning a persistent identifier.

Accessible

Accessibility depends crucially on the selection of a suitable repository, i.e. a publication platform with the possibility and guarantee of long-term archiving, including the publication options. The data formats and protocols used should also be open, free and universally implementable. The metadata should be accessible, regardless of whether or not the data themselves are available.

Tip: re3data can help you find a suitable publication venue or repository. In general, there are three types of repositories: ‘institutional’, ‘subject-specific’ and ‘generic’.

Interoperable

 Interoperability depends a lot on the file formats and the data-naming conventions. A truly open file format is just as desirable as open software, provided the long-term availabilty of the latter is ensured. When considering the options, particular attention should also be paid to file formats that are widely or fundamentally used in the (specific) scientific community. The file structure should be standardised in terms of naming conventions, following the principle of ‘short but meaningful’ and avoiding certain characters such as <> # ! ;

Reusable

Data can be reused by granting an appropriate licence. Creative Commons (CC) licences are widely used, but it should be noted that not all of them allow for Open Access use.

* According to the specifications of the German Research Foundation (DFG) and its guidelines, a dataset is considered to be long-term archived if it can be stored loss-free for more than 10 years.