Focusing on the generalization of concepts, functionality, and overall processes involved in the creation of a secure 'network of trusted data' , the IDS-RAM resides at a higher abstraction level than common architecture models of concrete software solutions do. The document provides an overview and dedicated architecture specifications.
@ssteinbuss thanks a lot for the clarifications! Here's what I captured from our discussion, please correct me if I got something wrong:
Look in the Process layer: #208: vocabulary service in the process layer
Plenty of these "Semantization" aspects are already covered
in particular, how Data Providers can make Data Offerings that include Self-Describing data
But there's no mechanism yet for Data Providers to collaborate to ensure that their Data Offerings are harmonized.
This is an aspect of Data Space governance, eg to mandate they use CSVW for describing CSVs, particular ontologies, SHACL shapes, etc
Can data spaces use Linked Data principles as alternative to Connectors (transferring data inbulk)?
LD principles are not in RAM 4.0, but can be in the Whitepaper, and make it into RAM 5.0
Following up on this discussion:
_Originally posted by @VladimirAlexiev in https://github.com/International-Data-Spaces-Association/IDS-RAM_4_0/issues/236#issuecomment-1429428675_
We should add the Linked data principles to the RAM and how to use them in data spaces, based on the semantic interoperability paper.