Knowledge bases and reference services

The next meeting of the Virtual Reference SIG (sponsored by METRO) will be on October 31 from 10 am to 12 noon. Margaret Smith (Physical Sciences Librarian at the Bobst Library at NYU) will lead the discussion with a presentation on how libraries are currently using knowledge bases to support reference services. For more details see the Virtual Reference SIG wiki page on this event.

There are a number of large issues that I hope our discussion might cover:

• Maintenance. How can you design a workflow that keeps older content accurate and fresh. Many URLs that you provide will be subject to link rot. Resources that are recommended may no longer be accessible or still the best ones to consult. Who will be authorized to add content? Just librarians? Librarians and patrons? Librarians, patrons, and the whole world?
• Privacy. If you are adding patron's questions (even highly edited versions), how can you be sure that a patron would not be annoyed upon finding his/her question publicly findable?
• Findability. How will librarians and patrons discover content in the knowledge base? Will there be metadata added to records? If so, what kinds of metadata would be helpful? Will search be the main way that users navigate or will browsing be available (or even privileged over search)?
• Content. What exactly could go in? Transactions from email reference? face-to-face reference? chat reference? reference consultations? FAQs? Saved searches from databases and the catalog? comments and reviews appended to items in the catalog (and even databases, were database vendors to allow such functionality?)
• Placement. How will you alert librarians and patrons about this resource? Should it be added to any list of subscription databases? Group it in with any federated search tool? Keep it set aside on the library web site in location where tutorials, help guides, ask a librarian services, etc. are featured?
• Nomenclature. What should such a tool be called? And as far a librarian jargon goes, is it better to follow the more popular convention of spelling it as two words ("knowledge base") or does one word("knowledgebase") read better and mimic the word "database" in useful ways?

I hope that I'll have a chance to bring in some of ideas expressed by David Lankes in his presentation at ALA Midwinter 2008, where he discussed an interesting vision of a tool for capturing knowledge, recording it, and making it findable (scroll down this OCLC page to the link to the January 12, 2008, event).