SUBSCRIBE NOW!

Vol. 12 No. 7 — July/August 2004

Fortunately, the detective remembered the sage advice his mother had given him as a small child whenever they saw such sad cases. Mother had warned him: (1) never let them get behind you, and (2) never look them in the eye. (When you think about it, such advice might leave the "advisee" looking a little odd too. The bodily contortions necessary to avoid eye contact while still positioning oneself vis-à-vis another's location could get tricky.) The one thing the detective noticed about today's one-sided conversationalists that differed from the wackos of his youth was the presence of wires dangling from, or devices pressed to, their ears.

Some of these odd ducks even have drivers' licenses. Look out your window the next time traffic stalls on the freeway and you may see the driver of the car next to you engaged in an animated discussion — gestures, laughter, pursed lips, furrowed brow, nods, etc. — and all in a vehicle with no passengers.

Most of these people think they're cool. (Some probably think they're Napoleon.) But one thing they definitely are, dear reader, are our clients. Yes! These strange, sad, vulnerable beings have been given into our charge to protect and keep from harm to the best of our abilities. If we fail, what will happen to these babes in the woods, lying — as Shakespeare put it — "naked to mine enemies"?

Trouble looms in the darkness ahead. And the danger comes mainly from the lights in the darkness. Life in the information profession has taught us all that "sometimes the light at the end of the tunnel is an oncoming train." This issue of Searcher is replete with wonderful articles, but the hidden dynamite lies in Amelia Kassel's article on the spring workshop of the Southern California Online Users Group (SCOUG) ["Text Mining for Reputations: SCOUG Spring Workshop, 2004," pp. 58-61]. Admittedly, as a lifelong member of SCOUG and a planner for this particular workshop, I have a prejudiced outlook. Nonetheless, the promise of the new text-mining platforms for building sophisticated answer products is only matched by the dangers in accepting those answers without vigorous monitoring.

One thing is clear: Such vigorous monitoring will not come from our end-user clients, even the most brilliant, most energetic, and most sophisticated. Such monitoring takes discipline, focus, persistence, and time. End users will lack the time, the will, or the talent. If you don't believe me, consider this statistic. For many years now, through several generations of Windows' development, Microsoft has been tying more and more of its feature improvements, and/or the quickest, simplest way of reaching key features in its programs, to the right-click button on the mouse. Yet Microsoft studies show that 75 percent of Windows users do NOT use the right-click button. Windows' market dominance means practically everyone with a computer uses Windows to do practically everything they do on a computer. Yet only a quarter of them take the trouble to use a second finger in tapping the potential of the software they use. Not one more finger!

Now answer products are on their way. According to a squib in PC Magazine, Microsoft itself is only 2 years away from releasing its next-generation competition to Google, called Answerbot. For an early version of where Microsoft is heading, you might look at Microsoft's Newsbot and Blogbot services, which may be out this year. These new products will not proffer sites where information might appear, as Google and Yahoo! Search and other Web search engines do now. These products will simply answer questions authoritatively — and briefly, if people plan to receive their answers on tiny PDAs. Most likely the answers will come with an option to check the source. Perhaps all users will have to do is right-click, hmm? We could end up looking back on these Google-searching days as a golden era of Web scholarship, where a "greatest generation" of Web users actually took the time to read URLs and select sources.

Nor are the answer products limited to quick, short, factual, "look-up" answers. Reputation monitoring services are the first generation of products coming to commercial release from these tools. Factiva's Reputation Manager, done in partnership with IBM Web Fountain's top-of-the-line text-mining platform, could wipe out half the staff at Fortune 500 competitive intelligence centers. Its six-figure annual subscription price might help motivate such staff reductions. The Factiva service is very sophisticated, building evaluative factors into its assessments. This kind of product intends to achieve the status of an information adviser, not just an information source.

We are moving closer to the realization of the science-fiction dream of computers that can protect one from harm on their own initiative, ones so knowledgeable and nurturing that Will Robinson can sleep in peace until he hears them call. But is it safe to sleep in their digital arms? The sense of safety and security has such appeal that it could lure users into believing without checking, into trusting without monitoring, into acting upon information without verifying its accuracy.

Who shall save the innocents? Super Searchers to the rescue! The information professionals serving users must rise to the task of assessing these new services, of separating the wheat from the chaff, of monitoring when the wheat services start switching to the cheaper chaff. Information professionals must also identify new answer product services that could solve client problems effectively and efficiently and promote their adoption vigorously.

How shall we perform this task? First step, we must understand not just the questions, but the questioners. We must know what people will ask and how, what answers they will accept and why, what decisions they will make based on those answers, and whether those answers warrant those decisions. As always, first comes the reference interview, only now, instead of an individual interview with one requester and one request, it becomes an ongoing collection of data on our client base (including the people who could use us but don't).

By the way, the shape and structure of building such a client database is the subject of SCOUG's annual retreat, held in Santa Barbara at the end of July [http://www.scougweb.org]. If you can't join us there, Searcher will report the results of the meeting in a future issue. And readers can expect to see more and more coverage of the challenges coming from new answer product platforms in future issues.

...bq