The Technical Committee and Liaison Group of the OAI Object Reuse and Exchange effort met last week in New York City to hammer out face-to-face some of the last remaining issues before work could begin on a draft specifications document. In preparation for that meeting, we worked on a white paper that is officially called "Compound Information Objects: The OAI-ORE Perspective" -- but it might more accurately be called "Thoughts on Compound Documents". The outline of the white paper looks like this:
- Introduction and Motivation
- An OAI-ORE Interoperability Layer for Compound Objects
- Exposing Logical Boundaries in the Web Graph
- Named Graph Publishing
- Named Graph Authorship
- Generalizing Named Graph Publishing
- Named Graphs for Compound Objects: Issues for Exploration
- Connectedness and the Containment Node
- Publishing Named Graphs versus Publishing RDF Documents
- Discovery of Named Graphs
Sections 1-6 set up our collective thoughts on compound documents in which we borrow the concept of "named graphs" from the semantic web community. One of the things learned over our two-day face-to-face meeting was that "Named Graphs" in the semantic web community means a lot more than the "named graphs" term we use in this document, so a) don't read too much into the use of our term "Named Graph"; and b) the term is likely to change as the specifications are drafted to avoid confusion with the semantic web community. Section 7 sets up the issues that were identified in conference calls and e-mail with the scheme outlined in the previous sections. In the face-to-face meeting we came to some conclusion on all of them except Versioning, and I'll try to give insight as to the decisions made last week in a future post.
We (the members of the Technical Committee and the Liaison Group) felt comfortable enough with the document to publish it on the open web and solicit comments. Feel free to leave comments here or contact me directly; all comments will be forwarded back into the Technical Committee.