The LC BIBFRAME team recently has had the opportunity to take a hard look at annotations, and we would like to share our current/tentative thinking.
Some background: when we first started thinking about BIBFRAME annotations (more than two years ago) we thought seriously about whether we could base BIBFRAME annotations on the Open Annotation Model. After a brief effort to model annotations as such, we abandoned that approach, because we felt that our annotation requirements were too specialized to fit within the general model. Long story short and jumping to the present, we no longer believe this to be true. The earlier OAM community work has evolved and moved to the W3C (as “Web Annotations”, a Working Group has been formed chaired by Rob, and I am an active member) and we plan to again examine to what extent we can leverage this work for our benefit.
Let’s look at the BIBFRAME Annotation classes one by one and see how we might proceed.
Review and Description: I am reasonably confident that these types of annotations will be supported by the W3C spec, and therefore we will not need explicit BIBFRAME types.
CoverArt: there is also a reasonable chance that cover art will be accepted by the W3C group.
TableOfContents: Table of contents is a problem. There is little chance that it would be accepted as a W3C use case. We should discuss this in a separate thread.
Holdings: We are seriously considering suggestions to abandon the attempt to model holdings as annotations, and to instead define bf:HeldMaterial as a subclass of bf:Resource (and bf:HeldItem a subclass of bf:HeldMaterial). It is likely that this is the direction we will take.