Access permissions

Before accessing the Human Brain Project/EBRAINS Knowledge Graph through fairgraph, you must read and accept the Terms of Use, and then e-mail support@ebrains.eu to request access.

Public and private spaces, released and unreleased nodes

Each metadata node is stored in the KG in a “space”. Each space has its own access permissions. Some spaces are public, others private.

The KG also has the concept of released and un-released metadata nodes. Released nodes in public spaces can be viewed by anyone through the KG Search UI, without an account.

For programmatic access, for example through fairgraph, an EBRAINS account is needed to access metadata.

Anyone with an EBRAINS account can access released nodes in public spaces. Access to unreleased nodes in public spaces is restricted to curators.

To see a list of spaces you can access, run:

>>> from fairgraph import KGClient
>>> client = KGClient()

>>> client.spaces(names_only=True)
['collab-my-project',
 'collab-another-project',
 'common',
 'computation',
 'controlled',
 'dataset',
 'files',
 'in-depth',
 'kg-search',
 'livepapers',
 'metadatamodel',
 'model',
 'myspace',
 'review',
 'software',
 'spatial',
 'tutorial',
 'webservice']

To see list of spaces for which you have write access:

>>> client.spaces(names_only=True, permissions=["WRITE"])
['collab-my-project',
 'collab-another-project',
 'myspace']

For a full list of available permission names, run:

from fairgraph.client import AVAILABLE_PERMISSIONS
print(AVAILABLE_PERMISSIONS)

Working with private spaces

Everyone with an EBRAINS account can create metadata nodes in a KG private space named “myspace”. Only the user who created the node can access it.

To collaborate with others, users can create a workspace (called a “collab”) in the EBRAINS Collaboratory, choose who can access that workspace, and then create a private KG space for that workspace.

Before you can use a private space, it needs to be configured, by specifying which metadata types you plan to store in it. This is done with the KGClient.configure_space() method:

from fairgraph.openminds.core import Dataset, DatasetVersion, Software, SoftwareVersion, Person

types = [Dataset, DatasetVersion, Software, SoftwareVersion, Person]
space_name = "collab-my-project"
client.configure_space(space_name, types)

Now we can save a node to this space:

>>> person = Person(given_name="Oliver", family_name="Hardy")
>>> person.save(client, space="collab-my-project")

To see a list of how many instances you have stored in the space:

>>> client.space_info("collab-my-project", release_status="in progress")
{fairgraph.openminds.core.products.dataset.Dataset: 0,
 fairgraph.openminds.core.product.dataset_version.DatasetVersion: 0,
 fairgraph.openminds.core.actors.person.Person: 1,
 fairgraph.openminds.core.product.software.Software: 0,
 fairgraph.openminds.core.product.software_version.SoftwareVersion: 0}

If you want to remove all nodes from a private space, you can run:

>>> client.clean_space("collab-my-project")

Warning

Be very careful when doing this; it is really only intended for use when developing a new script that writes to the space, so that you can clean up mistakes before re-running the script.

Data Protection

fairgraph is a client library that communicates with the EBRAINS Knowledge Graph API on behalf of the user. It does not store, log, or process personal data independently. Any metadata transmitted to or retrieved from the Knowledge Graph — including researcher names, affiliations, and contact details embedded in research products — is subject to the EBRAINS Privacy Policy and the Terms of Use.

Authentication tokens passed to KGClient are used solely to authenticate requests to the EBRAINS API and are never written to disk or transmitted to any third party by fairgraph.

For questions about GDPR compliance of the EBRAINS infrastructure, contact the EBRAINS support team via the EBRAINS helpdesk.