FAIR - Findable

Revision as of 19:38, 25 July 2021 by P.petrelli (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Data should be easy to find and identify

Publish in trusted repository

The best way to make data findable is to publish it in a trusted repository. If possible, choose a repository which is discipline specific and used by your community. Think of places where you go to look for data you want to use for your own research. A good repository should care about the quality of the information you attach to the data and how your data is formatted. The best repositories have some data requirements you need to satisfy. You should not see these as meaningless obstacles but rather as an opportunity to improve your data presentation. Usually these requirements are very reasonable and after the first time you submit some data you will be familiar with them and the process will be faster. It is possible to list your data in more than one repository; this is perfectly acceptable and encouraged to increase the data visibility. However, be careful to make sure you always have only one official record with an attached DOI, to which all the other listings refer to.  

Use identifiers

A Digital Object Identifier (DOI) provides a globally unique and persistent identifier that can be resolved online, they have been used for papers for years but can now be applied to data. A DOI allows users of your data to easily reference it and for other researchers to discover and access it from the reference.

Whenever possible, use other available identifiers such as grant, researcher and projects identifiers. If you keep all your research products connected in this way, anytime a researcher looks at one of your research product, they are more likely to discover the rest of your research.  

Rich metadata description

If you have your data on a repository, with an attached DOI but your metadata is poor, it would be hard for a researcher to actually find your data among many other datasets. Services indexing data use keywords, tags, subjects, the dataset title and main abstract to return results to a query. You should try to use all of these and provide accurate and detailed metadata when publishing your data to maximise its discovery potential. Conventions and standards help defining what kind of information should be covered.


Related pages

Persistent research identifiers

Climate science conventions and standards

Data management plans

Readme files