How are research data described in RADAR?
The description of datasets is one of the tasks of the (sub)curators. RADAR has developed a metadata schema for the description of research data that specifies 10 mandatory fields. The mandatory fields meet the requirements for the assignment of a DOI (Digital Object Identifier) for the dataset. The fields can be used across different scientific disciplines. In addition to the mandatory fields, 13 optional fields can be filled in. Thorough descriptions increase the visibility, findability and reusability of the dataset. The documentation of the RADAR metadata schema is available here.
Are the dataset and metadata record checked for quality by RADAR?
The technical integrity of the dataset is checked by RADAR at regular intervals in the process of bitstream preservation. Checking the content of dataset is the responsibility of the data providers.
The metadata is automatically checked for completeness and validated before publication or archiving in RADAR.
Which databases and search engines can be used to find datasets published with RADAR?
RADAR uses the protocol OAI-PMH to offer metadata for automatic harvesting by search engines. The metadata is licensed under a CC0 license and can also be accessed through the DataCite metadata store. These measures ensure that the published datasets can be found through a variety of search engines and databases.
Can metadata be added automatically?
Yes, administrators and curators can compile XML-files containing metadata offline and upload them to RADAR. In this way, datasets can either be described individually or by adding default values that were predefined for an entire workspace. A template for XML-files is available online in RADAR. Metadata can also be uploaded automatically using the RADAR API. First steps for using the API can be found here.
How is the interoperability of metadata ensured?
The RADAR metadata schema is compatible with the widely used DataCite and Dublin Core metadata schemas. In addition, the RADAR metadata can be found and reused via the DataCite metadata store. A connection to ORCID is implemented.
In which languages can metadata records be entered?
The language for the metadata is chosen by the data provider. A translation of the metadata into several languages is not yet planned. Optionally, a language for the dataset (i.e. the resource itself) can be specified.
Do licenses refer to the entire dataset or to individual files?
The license specified by the data provider refers to the entire dataset. If data providers want to specify different license terms for individual files, the data must be split into several dataset. RADAR recommends the use of Creative Commons licenses.