Open JohannesLipp opened 3 years ago
Comment by @clange : the idsc:ANONYMIZE that we see here should definitely be aligned with DPV. ODRL's notion of "action" (which IDS reuses) looks really the same as what DPV calls "process".
So, DPV defines "Processing" as "The processing performed on personal data", and ODRL says "An Action class indicates an operation that can be exercised on an Asset." And now we know that "anonymize" is an action/process that only makes sense on personal data, so i think we should align these two for the purposes of IDS. I don't see DPV aligning itself with ODRL so far, but I will ask that question to the DPV people.
Please also note https://github.com/dpvcg/dpv/issues/11
Hi! Just to let you know that I have already been doing work on aligning ODRL and DPV (online documentation), so if this topic is still of interest to you, I would be more than happy to discuss as I am also starting to look at policies in data spaces :)
Please align the IDS policy language (where possible) with subclasses of dpv:Processing.
(See https://dpvcg.github.io/dpv/#vocab-processing-categories – there is, e.g., https://dpvcg.github.io/dpv/#Anonymise and https://dpvcg.github.io/dpv/#PseudoAnonymise. But, OK, strictly spoken, these are not, like in IDS, states, but operations ("The processing performed on personal data"), which lead into a state. Nevertheless, aligning (at the very least by rdfs:seeAlso) makes sense. (Issue proposed by @clange )