Ed Summers has a nice note on the Worldcat Registry of institutional profiles. Talking about machine level access through web services.
The details are available here (free worldcat.org registration required). Two web services are now available. There are full details on the pages; here is what the headlines say:
WorldCat Registry Search is a Web service that retrieves a set of "thin" records for Registry-profiled institutions or consortia that match specified criteria. The criteria could be any combination of institution name, alias, type, city, state, country, postal code, OCLC Symbol or RLG ID. The service returns a set of records in XML format that meet those criteria. The records returned by this Web service are the same information normally displayed to an unauthenticated user who conducts a search on the WorldCat Registry Web site.
Worldcat registry detail. This Web service retrieves details about a single institution or consortium profiled in the WorldCat Registry. The service uses the Representational State Transfer (REST, or "RESTful") software architecture and retrieves a record from the Registry based on a specified numeric WorldCat Institution Identifier (ID). Results are returned in XML.The former uses SRU. Here is an example of the latter.
Click on the following for the xml response:
and here is the same entry in the (humanly) web acessible registry:
Update: I meant to mention that the lead architect for the registry is my research colleague Jeff Young, working closely with product and development folks in Dublin and in Mountain View.