Currently, we are using simple npx:approvesOf pointing to agent intros.
In the future, we need dedicated "endorses"-relations to minimize the size of the nanopublications in the core. Moreover, these relations should be qualifiable with a community and/or domain identifier, so one can endorse somebody for a given community/domain, and thereby "approved" agents for a particular domain can be calculated.
This needs some more thinking about the general trust algorithms too...
Currently, we are using simple npx:approvesOf pointing to agent intros.
In the future, we need dedicated "endorses"-relations to minimize the size of the nanopublications in the core. Moreover, these relations should be qualifiable with a community and/or domain identifier, so one can endorse somebody for a given community/domain, and thereby "approved" agents for a particular domain can be calculated.
This needs some more thinking about the general trust algorithms too...