DOIs on datasets

The IDOC (Integrated Operation and Data Center) platform at IAS has just assigned DOIs for 83 numerical codes and datasets it hosts. A DOI (Digital Object Identifier) is a character string that uniquely and permanently identifies an electronic resource.

DOIs are commonly used to uniquely declare and identify scientific publications, but they can also be associated with computer programs or datasets. This is how they are now used by IDOC.

DOIs were created to overcome the obsolescence of URLs. With DOIs, you can find a digital resource even if it has been moved. Take the example of the second data release by the SPICE instrument of the Solar Orbiter mission: the raw form of its DOI is 10.48326/idoc.medoc.spice.2.0 The Web address (URL) corresponding to the DOI can be obtained by simply adding the prefix https://doi.org/. Thus, the address https://doi.org/10.48326/idoc.medoc.spice.2.0 will always allow you to locate this dataset, regardless of the evolution of its supports (infrastructure, environment, access software, etc.). The home page pointed to by the DOI must describe the resource and all the elements enabling its optimal use.

Several platforms are authorised to allocate these DOIs. CrossRef, mEDRA and DataCite are the main agencies or associations that have been responsible for issuing these permanent identifiers since the year 2000. It is the DataCite platform via the Institut de l'information scientifique et technique (INIST, CNRS) that delegates to IDOC the registration of our DOIs in return for an annually renewable contract.

The identification of a resource by its DOI also facilitates the feeding of other platforms or data warehouses. Its use is recommended by the Practical Guide of the CNRS Ethics Committee (COMETS). In the HAL database (CNRS), for example, all you have to do is enter the DOI of a document to add it. The repository then automatically imports the associated metadata (author, type of publication, publication title, date etc.)
 

Figure 1: Example of upload to HAL of a document identified by its DOI. The document metadata are then automatically imported.

In the same way, it is possible to import all the metadata of a document on ORCID (Open Researcher and Contributor ID) from its DOI, or to search scientific papers on the ADS database by their DOI.
 

Figure 2: Example of DOI import into ORCID

Today, 83 codes and datasets hosted by IDOC are identified by their DOI.