Use case for personal curation in the repository

[stonematt]

I'm trying to map my expectations of how we issue and validate professional credentials into the documented use cases, and I think we have a gap. UC C.4 https://opencreds.github.io/vc-use-cases/#professional-credentials implies what I'm thinking, but I'd like to express a use case that's centered around the individual's process for collecting and curating their own data.

This process would bridge several of these categories of use cases. In C.4 we state that Jessica "has a variety of digital claims" but none of the uses cases suggest how that came to be and interacting with the credential repository is pretty ambiguous. I'd like to add a use case in this area. A repository is likely to serve the "whole individual", storing claims that speak to a combination of education, forma professional credentials, professional training, legal identity, interests and hobbies, etc.

I'd like to see explicit use cases that deal with this tactical exercise of an individual managing their portfolio -- adding, removing, maybe even generating bundles of claims for sharing and validation.

Is this a new category? Anyone else think this would be a useful add?

[jandrieu]

I agree. After diving in with the prescription use case #38 I had a similar thought.

If we're on the same page, I'd suggest we document a few (not dozens, just a few) of these "porftolio management" exercises using Engagement Models, which I proposed in #34.

The Joram 1.0.0 Engagement Model I've been working on from Rebooting Web of Trust is probably a good example of that. https://docs.google.com/document/d/1GLejHAyOGcFZMDH23VpBK5as_474gt1tdYZIWkHm7c0/edit?usp=sharing

That model is almost entirely technology independent by design. It doesn't discuss how verifiable claims would be used. For our purposes, it would make sense to integrate verifiable claims in whatever engagement models we adopt.

[stonematt]

@jandrieu would a simple definition and description of the "wallet" be sufficient to handle this issue?

[jandrieu]

@stonematt Perhaps, although I worry a simple definition will either over-simplify or be too complex. We have at least seven things a wallet /might/ keep track of:

1. keys / authentication credentials
2. verifiable credentials aka attributes/data
3. presentations aka profiles/personas
4. consent records (both for the individual and presentation recipients as consenting to specific terms)
5. supported DID methods
6. proxy resolvers (method linked)
7. issuers (optionally)
8. policies (optionally)

2, 3, 4, and 7 could be handled either locally (in wallet), nearby (in a user's server or remote device), or remotely (in the "cloud" whether a hosted service or a distributed repository like IPFS). Supported methods is likely hard coded, but one could also imagine a way to dynamically load a method handler. Policies mean pre-set terms that the user might attach to presentations or for which they auto-accept (or semi-auto accept) a verifier's offer.

And these are just off the top of my head. There's been minimal standardization of wallets for DIDs and VCs, but we may as well get started.

[jandrieu]

https://github.com/WebOfTrustInfo/rwot7/blob/master/draft-documents/Digital%20Credential%20Wallet.md

[stonematt]

by Curate, I mean the actions available to the holder, as they choose an arbitrary collection of verifiable credentials that they collect into a Verifiable Presentation to present to a verifier.

[jandrieu]

@stonematt Would you like to advance this use case? We're open to new ones.

Here's the current call for input.
https://lists.w3.org/Archives/Public/public-vc-wg/2023Apr/0000.html

We'd welcome your contribution.

[jandrieu]

I reached out to @stonematt out of band. We'll mark this pending closed and see if he still is inspired to put something together.

[stonematt]

@jandrieu Thanks for the heads up. I need to take some time to get my head back in to this space. The last comment was (unbelievably) >5y ago and the group has marched forward on capabilities. Don't close it quite yet. I'd like to cycle on it.

[jandrieu]

Great. We welcome the discussion.