FAQ section

[pablo-de-andres]

Present the most important concepts in a Frequently Asked Questions section to simplify understanding

[pablo-de-andres]

@yoavnash and @urbanmatthias provided some initial questions:

Once an ontology is loaded via PICO into the OSP-core, are the CUDS the only way to interact with it?

Once the ontology is loaded, the user can interact with it similar to python classes. There is a rdflib triplestore in the background, but it is not recommended to access that as a user.

Is the ontology only kept in such graph (osp_ontologies) format? Is it your internal, proprietary representation?

~/.osp_ontologies is a directory where OSP-core moves the installed ontologies. In there is a graph.xml file which contains all installed ontologies in OWL format (i.e. non-proproetary)

Wouldn’t it be also possible for the OSP-core to manage an ontology directly from a triplestore?

It kind of does. The ontology is loaded in an rdflib triplestore and OSP core provides an interface to access it similar to python classes.

Let’s say that the system is running and individuals are created in the way we discussed (so some CUDS objects are created). Is there a way to export classes + individuals from the OIE in owl, for example, or other formats?

Yes, there is. See from osp.core.utils import get_rdf_graph, which returns an RDF graph.

[pablo-de-andres]

Removed from contribute section as it might be more relevant in the introduction