FRIR
Functional Requirements for Information Resources is an extension of FRBR to apply to digital [data files].
Its namespace is defined at http://prefix.cc/frir
Diagram of the FRBR ontology: pdf
The gist of FRIR is that every file is its own distinct frbr:Item, and above it is its own "stack" of frbr:Manifestation, frbr:Expression, and frbr:Work. BUT those stacks can be "pinched" together when they are the same at that level. frbr:Manifestations are identified using message digests (MD5, SHA), while identifying frbr:Expressions requires some assumptions/interpreting the structure of the file to "canonicalize" into a "content digest". frbr:Work is completely abstract and requires "human assertion" based on (often) out-of-band information.
frbr:mccusker2012parallel is the landing page for the “Parallel Identities” journal article that introduces FRIR, its pdf is available.
IPAW paper came after that, should cover same material but shorter and might have some more recent thoughts:
Implementations:
Capturing http://prefix.cc/htir during the HTTP request is part of the solution to remember the pingback URL when downloading a file, so that you can POST back a pointer to your derived downloaded file (as demonstrated in the Walking into the Future paper).
Jim's thesis adds in a better graph digest algorithm and digital signatures.