Comments on Authority Control Scenario

This concludes my comments on the Research Drivers document.

Nick Matsakis

4.3  OCLC Authority Control Service: I think it is important to make a
distinction here between a service that simply provides OCLC authority
records to SIMILE and a service which matches unauthoritative records to
authorities within SIMILE.  The scenario/context section here (4.3.2)
seems to be talking about simply providing access to the records while the
use case (4.3.4) seems to be focused on accessing a full-fledged authority
service.  My understanding is that OCLC provides such a service and if we
are discussing publishing _that_ service as a web service, that should be
made more explicit.

I'm also not sure whether the matter of differences of opinion on
"correct" metadata belongs here.  There is definitely authority work that
deals with ambiguity of that sort; for example the question of "Is this
work about that topic?"  However, it seems to me that most authority work
is concerned with the question of "Is this entity the same as that
entity?"  which generally has a simple answer --- though finding it can be
difficult when there is not enough information about the entities to make
an identification.

Instead, I would put the focus squarely on providing authority service in
SIMILE/DSpace's heterogenous data environment. I am not familiar with the
details of OCLC's authority service, but I suspect that it only works with
MARC metadata.  I expect that the more flexible DSpace environment will
bring with it challenges that these services aren't currently capable of
handling. An example of a substantial body of metadata that isn't
trivially translatable to MARC would be a good argument for  for providing
a new authority service that is able to work with more generic metadata.
Otherwise, Schema translation services may enable existing MARC-based
authority services to work with non-MARC metadata.

Received on Thursday, 10 April 2003 15:19:26 UTC