OK Brian Lowe needs to breathe a couple of times if I am going to have *any* chance of catching what he is saying. He’s talking about VIVO, which is an identifer system primarily funded by the NIH in the US.
VIVO focused on a linked data approach – one URI: public profiles for humans with data for machines. The URIs are assigned by the institution providing the VIVO instance and are structure to served linked data but no semantics are imposed…although it is assumed that it will be some part of an institutional domain space, e.g.: http://vivo.cornell.edu/somestuff.
VIVO assumes there will definitely be multiple URIs for people as we move institutions. This immediately asks the question…why the affiliation approach and what value does this add??
VIVO has a core ontology that focuses on mapping people with organizations using existing frameworks where they exist…e.g. FOAF and BIBO. This can then be extended locally with institution specific semantics. VIVO includes information not only about how stuff is related but also when….assuming that time context is specific and relevant to the author identifier space. They focus on moving beyond the publisher space arguing that publishing is just one instance of the application of an author identifier. Is this all approach the right one? I’m sure the approach will come up for debate. VIVO also relies quite a lot on local systems to populate…which gets back to the consent question and how this should be managed.
Question from the floor on who normally takes the role of supporting the VIVO node – answer is in differs. This is a fairly typical response to IDM approaches and one we are familiar with in the REFEDS space.