RADAR has always supported the FAIR principles to make RADAR datasets as findable, accessible, interoperable and re-usable as possible. With the current RADAR software release, we have significantly further optimised RADAR's FAIRness:
On the one hand, we have updated the RADAR metadata schema (v9.1) and thus further developed the possibilities of metadata annotation.
On the other hand, we have optimised the RADAR landing pages according to the signposting approach, thus implementing an important part of the Fair Digital Object (FDO) framework.
The improved RADAR software version v.1.15 will be available from 14.12.2022.
We hope to have optimised the metadata annotation options and especially the FAIRness of RADAR datasets in a future-oriented way and in your interest!
New RADAR Metadata Schema v9.1
The RADAR metadata schema is cross disciplinary, based on the DataCite metadata schema and contains ten mandatory fields and 13 optional parameters. It offers a combination of controlled lists and free text fields. Previously, standards data for individuals (ORCID ID) and funding organisations (CrossRef Open Funder Registry) were already supported.
In the updated metadata schema v9.1, we have integrated two more important standards data via application programming interfaces (APIs): Standards Data for Institutions (ROR ID) and the Integrated Authority File (Gemeinsame Normdatei, GND). We have also further extended existing controlled vocabularies and made outstanding adjustments to the current DataCite metadata schema (4.4).
The changes in detail:
In the metadata fields creator, contributor, publisher, rights holder and funding, institution names can be selected using the Research Organisation Registry (ROR), the global registry of open, persistent identifiers for research organisations maintained by the research community. In this way, researchers can ensure that their information on organisations or organisational affiliations is always unique, standardised, persistent and machine-readable. ROR is integrated into RADAR via the ROR API. As soon as users start typing the institution name, a suggestion list is displayed from which the desired entry can be conveniently selected. The organisation name is then permanently linked on the RADAR landing page to the persistent entry in the ROR register, where further up-to-date information on the corresponding organisation can be found. The standardised ROR ID is also stored in the metadata. (Alternatively, institution names can still be entered via free text. A link to ROR at landing page and metadata level is omitted here).
In the metadata field keyword, researchers can now also choose standard data records from the Integrated Authority File (Gemeinsame Normdatei, GND) as an alternative to entering free text. The GND currently contains more than 9 million standard records for persons, corporate bodies, congresses, geography, subject headings and work titles. The integration into RADAR, which takes place via the LOD interface lobid-gnd of the Hochschulrechenzentrums NRW, meets the needs of many humanities communities in particular (but not only!). As soon as users start typing the keyword, a suggestion list from the GND is displayed, from which the appropriate standards data set can be conveniently selected. This is permanently linked to the persistent GND standards record on the landing page. The values of all attributes (keywordScheme, schemeURI, valueURI, classification code) are also stored in the metadata.
On the recommendation of our user advisory board, we have added 13 more common licence types for research data and software to the selection list for re-use rights in the Licence metadata field for the dataset: "Public Domain Mark 1.0", "Public Domain Dedication and License (PDDL)", "Attribution License (ODC-By)", "Open Database License (ODC-ODbL)", "Apache License 2. 0", "Common Development and Distribution License 1.0", "Eclipse Public License 1.0", "Eclipse Public License 2.0", "GNU General Public License v3.0 only", "GNU Lesser General Public v3.0 only", "BSD 2-Clause Simplified License", "BSD 3-Clause New or Revised License", "MIT License".
In addition, we have enabled linking to ePICs as related resources and added the new type "Survey" in the metadata field data source. Both additions were suggested by our user advisory board.
Finally, we have made outstanding adjustments to the current DataCite metadata schema (4.4) and added corresponding controlled values. These include, for example, further relations for related resources ("isDescribedBy", "Describes", "HasVersion", "IsVersionOf", "IsPublishedIn", "IsRequiredBy", "Requires", "IsObsoleteBy", "Obsoletes").
All changes introduced with RADAR Metadata Schema v9.1 are now available via our form-based metadata editor, in the sample files for uploading annotations via XML file, and via our RADAR REST API*.
* When using the RADAR API, please note: due to the update of the metadata schema, please replace "version": "09" by "version": "9.1" in the code block "POST body for creating a new dataset".
The signposting community aims to make the scientific web more machine-friendly. FAIR Signposting is currently not a formal standardisation process, but follows a low-threshold approach that enables machines to interact with science portals in a uniform way, bypassing proprietary APIs. The approach was discussed at the 1st International Conference on FAIR Digital Objects (26-28.10.2022, Leiden/NL), among others, and was considered promising for significantly promoting the interoperability of scientific objects on data repositories, publishing platforms, etc. and implementing the FAIR Digital Objects (FDO) Specification in a research data repository.
FAIR Signposting is based entirely on widely used web protocols specified in IETF RFCs. Patterns that occur repeatedly in scientific portals (e.g. DOI, author PIDs, web addresses of landing pages and descriptive metadata) are clarified via the inclusion of typed links. These typed links are provided in HTML resources in simple HTTP link headers and HTML link elements. Machine agents thus receive uniform and lightweight information about these elements and their web addresses and can navigate to them.
In line with the FAIR signposting approach, we have enriched the source code of all landing pages of RADAR datasets with typed links, thus improving their machine readability and actionability in a state-of-the-art way.