Some thoughts on Federated Search

OCLC’s 2005 report “Perceptions of Libraries and Information Resources” told us a dirty little secret: only one percent of information seekers start with the library web site, preferring easier-to-use web sites, even if the latter don’t lead to information of comparable quality. (from Casey/Stephens, The Technology Storm, Library Journal)

While that may be a dismaying statistic, it’s not, probably, a big surprise to any of us. It’s a sad truth — I sure know how many of our students haven’t even looked at library resources when they sit down despairingly at the reference desk. I also know how many are pleasantly surprised, even stunned, when they see the cool and perfect resources they can find in library databases. So why is it all of that true? Why do they not come to us first?

Part of it is an obvious education and PR problem about the role of libraries in the life of the modern student or user, which is an entirely other post.

Part of it has to do with the way we present our resources. Nearly every reference interaction — even the ones where the student says “COOL! That’s awesome, it’s just what I need, thank you!” — ends with me starting over from the library homepage and showing them again how I got to the database, how I did the search I just demoed, and instruction on how to do this at home, yourself. Knowing that the vast majority of our users don’t come to us for help, and instead will turn to the much clearer and easier (though less effective) Google if they’re frustrated by the library’s resources, I see that as an insurmountable barrier to performing successful research @ Your Library.

There was a lot of talk at Internet Librarian this year (and at Computers in Libraries, as well, and on the blogosphere and in the literature) about how our library-developed and library-vendor-managed resources plainly and frankly suck in contrast to the commercially produced and communally developed resources currently available to the people who we hope would be our users. Our interfaces? Are terrible in comparison. They’re clunky, they’re ineffective, they’re visually unappealing, they’re hard to use, they’re laden with jargon (“boolean”, “OPAC”, and “adjacency”, anyone?), and they are well behind the emerging technology curve. In short, the resources we’re offering are being completely outpaced and marginalized by our online competitors. As a profession of information consumers (yes, we are consumers — we pay for these things), we have to step up and demand better from our vendors.

So, presuming we demand such, while our vendors develop that perfect application that they will then charge us unholy prices for, what can we do? How do we work to close the gap in the meantime?

I think that federated searching is one option.

I’m a librarian. I understand why multifaceted searches and controlled vocabularies are to the benefit of the searcher, and I understand why the interface and presentation of results are less important than the quality of the information retrieved. I’m also a user of the web, and I know what I like, and I assume that some of my preferences transfer to other users of the web.

  • I like clean interfaces.
  • I like “current-looking” web pages.
  • I like easy search interfaces with an option to go deeper.
  • I like customization.
  • I’m comfortable sorting through huge result sets to find the best information.
  • I like having links that “go out” to more related information.
  • I like having the skills I learn in one interface be directly applicable to my next information need.
  • I like the One Stop Shopping model of web searching.

Knowing all of those things, I can’t help but turn to federated searching. Federated search products can do many, if not all, of those things while allowing the user to reach the important and relevant content in our catalogs and databases. I don’t think federated search is ever the option for all users — but for some…

For example:

  1. An undergraduate student wants “information about racial discrimination in the workplace” for a short paper for a COMM 101 class. Should I tell him that he’ll want to search four or five different “general use” databases, each with a different interface and quirks, or can I direct him to a single search box that will simultaneously pull results from our book catalog, Academic Search Premier, Opposing Viewpoints, and the Gale Virtual Reference Library? If we used a single federated search, I could also explain to the student that the next time he has a paper to do, on, say, the portrayal of male relationships in buddy movies, that he can use the same search process, because I’ve just taught him transferable skills.
  2. A graduate student is trying to do research on counseling methods in secondary education. Do I tell her that her first six steps should be to search the library’s catalog, then ERIC, then Education Research Complete, then the Proquest Education Journals, then Gale Virtual Reference Library’s education titles, then the Oxford Reference Online education titles? Or could I first direct her to a single search of our major online education resources through a customized federated search implementation? And then I could then follow-up with a bit of conversation about how to focus her search onto an individual database, using that interface’s strengths, if the federated results showed one source to be more appropriate than another for her topic.

There are many more iterations of how a federated search tool might reduce student frustration, increase research effectiveness, and generally provide a more current and responsive service to student researchers. There are equally compelling reasons why a federated search isn’t always the right idea. For example,

  1. Our psychology majors, who are taught through their entire academic program to use PsycInfo’s controlled vocabulary and specialized search capabilities, would not be well served by federated search when doing junior-year research. They want and need PsycInfo’s capabilities at that point. Ditto Sociology, Geology, and several other areas in which upper-division research is best served with subject resources.
  2. Our chemistry students, who rely almost solely on our ACS database, would not be well-served by federated search at any point in their chemistry program, because of the disciplinary publishing monopoly on the content relevant to them. That’s a meta-problem that meta-searching can’t address.
  3. Faculty researchers who know exactly which journals they need to consult to stay abreast of their specialties will not need federated search, as the A-Z list periodical titles provided by Serials Solutions will will serve them far more quickly and effectively in locating those titles.

But our History students might be well-served by federated search, since our library resources lack a one-stop-shop database for ‘history’ — would they benefit from having the catalog, JSTOR, Project Muse, the Alexander Street databases, and Academic Search Premier in one customized interface? And what about our interdisciplinary programs, like Women’s and Gender Studies, or Environmental Studies, or programs like Politics and Sociology whose research bleeds across databases — would a general federated search interface improve their research efficiency? How about our endless supply of incoming undergraduates doing general education work, taking a speech class, or a freshman writing course, or a critical thinking class whose topic might be anything under the sun — would we, by teaching those students early in their academic career to use a single interface that brought back broad but relevant results, be increasing the value of those GenEd courses? I see a lot of possibility.

Federated search isn’t for every search, or every user, every time. But I believe that by unifying our resources under one branded, simplified, and friendly search portal, federated search has the possibility of serving all of our students at some point in their academic career. And maybe, as a happy side effect, a federated search product could make the libraries seem more relevant, more up-to-date, and more accessible to our users. If we are those things, perhaps they’ll start to think of the libraries as a place to start their research, not to finish it.

—————-
Now playing: Tori Amos – Enjoy The Silence
via FoxyTunes

One thought on “Some thoughts on Federated Search

  1. Pingback: Interesting blog post re federated search in libraries » Federated Search Blog

Leave a Reply

%d bloggers like this: