Usage

The openMINDS schemas should be serialized using the JSON-LD format. For an automated digestion of openMINDS conform metadata into the Knowledge Graph (KG) database, the generated JSON-LD files should be organized in an openMINDS metadata collection and delivered as zip-archive to the HBP/EBRAINS curation team ([email protected]).

Expected JSON-LD syntax for storing openMINDS compliant metadata

An individual JSON-LD openMINDS metadata file should be conform with the following generic syntax:

{
  "@id": "<<openMINDS_JSON-Schema_@id>>",
  "@type": "<<openMINDS_JSON-Schema_@type>>",
  "<<property-label>>": "<<entry-value>>",
  "<<property-label>>": [
    {"@id": "<<openMINDS_JSON-Schema_@id>>"}
  ]
}

The expected entry value of "@id" and "@type", as well as the expected property labels and entry values are defined in each openMINDS JSON-Schema (see Implementation).

Structure of a KG database compliant openMINDS metadata collection

In a valid openMINDS metadata collection, the generated JSON-LD metadata files should be organized using the corresponding "@id" as file paths. For example, a JSON-LD file generated according to the v1: MINDS specifications in dataset.schema.json has the "@id": "minds/core/dataset/v1.0.0/dataset-01.json" and is therefore stored in an openMINDS metadata collection as follows:

....minds/
........core/
............dataset/
................v1.0.0/
....................dataset-01.json

Note that all JSON-LD files in an openMINDS metadata collection need to be conform with the correspondingly used openMINDS metadata schemas. Furthermore, each JSON-LD file should be linked to at least one other JSON-LD file in the same openMINDS metadata collection by referencing to its corresponding "@id". For this reason, each JSON-LD file within an openMINDS metadata collection has to have a unique "@id". In case of a locally defined openMINDS metadata collection, the unique part of the "@id" is established by the filename of the corresponding JSON-LD. In the KG database each of these locally unique identifiers is changed into a globally unique "@id" by replacing the local identifier of the JSON-LD with a universally unique identifier (UUID).

Practical examples for valid openMINDS metadata collections:
v1: MINDS | example-01

Making use of already existing metadata structures in the KG database

In order to avoid duplication of already existing metadata structures in the KG database it is also possible to directly make use of those instances. Especially for metadata structures that are commonly reused in various data descriptions, the KG database comprises a well maintained collection that does not have to be redefined for individual cases. A list with relevant "@id" of commonly used metadata in the KG can be found in KG Metadata Archive.