zschoenb
commented
6 years ago Finding: Jupiter is on the right path with respect to modeling HTTP(s) URI, as it contributes significantly to resource discovery by both human and machine. Modeling licenses and URIs is, for example, an entire recommendation in itself, and Jupiter has already accomplished this. Sitemaps are clearly important for discovery as well, and Jupiter has conformed to that need. We have pursued content aggregation through OAI, which means we are pursuing appropriate goals on that front.
Recommendation: Investigate implementing ResourceSync and, over the longer term, author entity modeling based heavily on ORCID. ResourceSync appears to be the highly preferred path forward for resource sharing, content aggregation, and batch discovery. Author entity modeling using ORCID appears to be the preferred path forward for enabling a globally networked system of tracking metrics, issuing push/pull event notifications, and integrating with third-party apps, including social media and recommender systems. Notifications will also require some metadata modeling using standards such as SUSHI or Linked Data Notifications. COAR recommends social web-integration, including annotation and comment systems. If Jupiter decided to go down this path, Activity Streams 2.0 and Web Annotation Framework would be necessary standards.
see the full report:
sfarnel
Review the COAR report (https://www.coar-repositories.org/activities/advocacy-leadership/working-group-next-generation-repositories/) with an eye to the metadata components.