I got a note from Eric Lease Morgan that ties together a couple of recent posts. Peter Murray noted the importance to libraries of managing metadata collections potentially pulled from various sources. And as I noted a couple of days ago, my colleagues have just written about the OCLC Research Publications Repository.
What Eric has done is harvest the metadata from our publications repository, as part of a larger collection of harvested metadata he is building.
To make this happen, I installed the latest version of MyLibrary on a machine in Oregon. I created three MyLibrary "subject" terms called "Library science", "technical reports", and "OCLC Research Publications". I then pointed my OAI harvester at your repository, collected the documents, and automatically "classified" them with each of the subject terms. Next, I wrote a report against MyLibrary listing all of the OCLC Research Publications and fed it to swish-e, an indexer. Finally, I wrote a simple CGI script providing access to the index. Sans installing MyLibrary, the whole process took about 30 minutes. [Personal communication, 3/16/05]You can look at the result of Eric's work here.