search

International-Data-Spaces-Association / IDSA-Rulebook

The working repository of the IDSA Rulebook Working Group
Creative Commons Attribution 4.0 International
15 stars 3 forks source link

Describe mandatory and optional aspect of semantic interoperability #4

Closed ssteinbuss closed 1 year ago

ssteinbuss commented 2 years ago

Describe mandatory and optional aspect of semantic interoperability

sebbader commented 2 years ago

"As a data space community, I want to declare certain vocabularies as mandatory in addition to the IDS IM."

Vocabularies = RDF vocabularies, OWL ontologies, leight-weight catalogs, industry catalogs like ECLASS, additional data models like AAS, ...

sebbader commented 2 years ago

"As a data space community, I don't want to be forced to follow domain-specific vocabularies that are declared by IDSA or similar top-level bodies."

sebbader commented 2 years ago

I just tried to put my view in user stories. To sum it up: I want to have the possibility to agree on mandatory vocabularies in my own data space but I don't want the IDSA to make this decision for me.

PeterKoen-MSFT commented 2 years ago

I would suggest that IDSA defines the way HOW vocabularies are treated if they are needed, but NOT WHAT vocabularies are going to be used.

My suggestion is that the self-description of the dataspace needs to contain information of whether there are any mandatory vocabularies in the dataspace and if so, where to find them and what the current version is, etc... This information is binding for anybody who wishes to join the dataspaces and therefor a pre-requisit to join the dataspace is going to be a proof to the adherence to this mandatory rules.

If a participant offers a data contract that requires a special vocabulary to be used in addition to (or potentially instead of???) the optional mandatory one, then this information would have to be again defined in the policies of the data contract (e.g. adding an obligation to use vocabulary xyz). The participant then would have to provide means (ideally through its self-description) to discover and access those additional vocabularies.

This way the dataspace authority has the autonomy to define which vocabularies, if any, are mandatory to be used in the dataspace and individual participants have the autonomy to define additional vocabularies if needed for the processing of the data contracts they offer.

ip312 commented 2 years ago

I would like to support Peter's proposal. In the MDS we are using both concepts.

  1. We defined additional mobility specific attributes for the description of data offers. One attribute is mandatory, so every participant has to use this attribute. This is binding, exactly as Peter has described it.

  2. As far as the data exchange itself is concerned, data from participants can be exchanged in any format. We will use the Vocabulary Hub to open the possibility to upload and download relevant vocabularies. Later on, it will also go towards harmonisation and matching between available vocabularies, the next step would be the setting up of an MDS-specific vocabulary, which will not be mandatory though.

anilturkmayali commented 1 year ago

An issue (that belongs) and to be discussed within the previous version of the IDS Rulebook. It's already addressed in the new version of IDS Rulebook.