What’s Next? Search Me
By Joseph Janes
American Libraries Columnist
Assistant Professor, Information School, University of Washington.
intlib@ischool.washington.edu
Column for December 2004
Ah, Dialog. Come back with me now and spend a moment basking in the halcyon glow of the days when the company exercised hegemony in the world of searching. (I myself never even learned BRS and always had to remind myself of the Lexis/Nexis commands, mentally translating them to “(2W)” and “/DE” anyway.)
In the introductory class at our library school, students build an inverted file by hand, on little bits of index cards, much as Dialog does. They then discuss how much power and control that kind of file gets you in searching. As the search world continues to evolve, the stability and precision that Dialog and other professional search tools stood for have become increasingly nostalgic notions.
Way, way back, in the early 1960s, individual database producers had to develop and provide their own search systems and interfaces. Dialog and its ilk provided common search command structures and made the searching world a good deal easier to work in, assuming you knew the commands.
Then along came the Internet. Early net searching tools (Archie, Veronica) were awfully crude, special-purpose beasts. Once the Web hit, though, and the earliest search engines arose, the little onscreen box became the standard search interface—and so it remains.
That little box is so obvious, so easy to use, that it doesn’t even require instructions (look hard at the Google front page). It is now pervasive—many library catalogs and bibliographic databases feature the little box, with advanced search features either hidden or demoted, emphasizing ease of use over power and sophistication.
Which makes perfect sense: For the vast majority of people, ease of use is paramount, and the relative few who want powerful advanced techniques will bore into the help pages to find them anyway. Why burden the casual, disinterested user with bells and whistles they won’t or can’t use?
Boxing match
From our professional perspective, the little box has not only diminished the potential sophistication of search, it’s also returned us to a more diverse search environment. Now all the major vendors (ProQuest, Wilson, EBSCO, OCLC, and so on) have their own idiosyncratic interfaces and features, and who can keep them all straight? Using these on the fly, with a reference client, is a real challenge.
Not to mention that normal people don’t get the differences between all these systems. After all, everything’s a little box, so what’s the difference between the library catalog and Academic Universe and Yahoo, anyway? An obvious response, federated searching (AL, Aug., p. 80), tries to search across a variety of resources and present a merged set of results; this is a mixed bag so far, but it’s certainly an idea worth serious pursuit.
In the Internet searching world, feature creep seems to be the Next Big Thing, as people try to distinguish themselves from Google (Google Print notwithstanding—I suppose I’ll have to write about that sometime, and the Google Desktop search, too). A9 (a9.com), from the people who brought us Amazon, has some keen toys and tools (and is an interesting place to start a shopping spree). The new and horribly named Clusty is Vivisimo with clustering capabilities and a Gossip tab. Icerocket is a metasearch engine with a bunch of goodies, as well as a much cooler name.
Then what? Feature development is one thing, and will likely continue in the short run. The world of search is now so competitive, and there’s so much money to be made in doing it well, that I can’t imagine some new approach or method won’t be introduced eventually.
What it will be, I can’t tell you. An article in the August 26 Economist described a prototype system out of Microsoft called Ask MSR that uses several natural-language processing techniques to manipulate questions (“When was Marilyn Monroe born?”) into sentences (“Marilyn Monroe was born”), which then can be searched. It’s more complicated—and difficult—than that; but the idea is simple and elegant enough, and just because AskJeeves never worked all that well doesn’t mean another question-answering system couldn’t succeed.
Even if it doesn’t, something else will. The world of search seems destined to be in flux for some time. It will get easier for everyone to do better searches with less preparation, less knowledge, and less technique, at least in the Web world. A boon to us, too, as well as yet further evidence that we must embrace the detailed, the complicated, the hidden, and the proprietary as our professional search domain . . . but that’s another story.
|
