How to refer to the data in a policy (IDS Artifact vs. ODRL Asset)

[hosseinzadeha]

When we specify a policy:

• Once we already have a set of documents that we would like to share on the platform. We can uniquely identify (a copy of) them and refer to them as an Asset Collection and specify a policy for their usage. In order to enforce the policy, the Asset Id would be sufficient.
• In another case, we are defining a contract for an Asset Collection that is not materialized yet; For example, we promise to transfer all bank transactions or all production plans of the following year. The policy rules shall stand for all those transactions/plans. When the data is produced and identified and during the Policy Enforcement process, we need to know that this data is an instance of that Asset Collection.

So far, in the IDS context, we would call a materialized data an IDS Artifact. A Policy is linked to an Artifact Id for the sake of simplicity in the process of Enforcement. However, we could link a policy to an Asset (or an Asset Collection) and be compliant with ODRL terminology, but also, we would like to imply that the policy stands for all the instances of those Assets.

Dear @riannella, now, here is my question: Does ODRL language assume that all Assets are materialized? Do we have any other term that would refer to a materialized instance of an Asset?

[riannella]

The ODRL model as always been relatively agnostic to the Asset identifier - it is the subject of the Policy rule.

If a system defines identifiers for assets/collections that don't exist yet, then the ODRL model will not prevent these being used in the expression of a Policy. The enforcement of the policy (which is not in scope of the model) would have to deal with an identifier that has no material asset.