Marydee Ojala Editor, ONLINE Magazine

the homepage Everything Old is New Again ONLINE, July 2001 Copyright © 2001 Information Today, Inc.

Developing effective search strategies has long been a professional concern for online searchers. In the early days of online, trainers stressed the power of Boolean logic. They explained how to visualize a research topic, find the facets that were important, and put them together into a building block, pearl growing, or successive fractions strategy. They taught us to look at descriptor terms. They taught us to order our thoughts. They taught us not to over-simplify, but not to over-complicate, either. They taught us the intricacies of formulating questions to retrieve relevant answers.

Then came the Web. Search strategy construction collapsed. Conceptualizing a topic disintegrated into "Enter your term in this search box." And that search box was much too small to accommodate sophisticated thought processes. Numerous end-user studies prove that they don't grasp the complexities of the true research process. End-users enter one or two words, regardless of the nature of their information need. They don't do a reference interview, not even with themselves.

The problem is that research issues are not diminished just because technology changes. Human thought processes remain complex. Peoples' ability to articulate their true research interests remains problematic. One person overstates the need; the next understates it: "Tell me about XYZ Corporation." "I want to know the smallest possible detail of the new product, unannounced, that XYZ Corporation is rumored to be manufacturing." Both want to know the same thing; they just can't state it properly.

Old-fashioned, yet common-sense, techniques apply just as well to Web searching as to traditional online. If you're in a space that contains only one topic, don't enter that as a search term. The classic example was, in the days of traditional online training, to avoid welding as a search element in the database Weldasearch (still online with Questel•Orbit, Fiz Karlsruhe, and Cambridge Scientific Abstracts). Similarly, if you've drilled down to a very specific level of Yahoo!'s hierarchy–say, travel in California–it's redundant and counterproductive to enter either travel or California. On the other hand, if you're only at travel in the Western States, you'll probably need to add the California concept

. Knowing which vertical universe you're in also alleviates ambiguity. Context is clearer. If you've restricted your universe to aviation, fly does not mean an insect. If you're in the petroleum universe, Shell is likely to be a company, not something found on the beach.

Not everything transfers. The open Web search engines are not as literal as the traditional ones. Command language searching, so popular at the beginning of online, has serious limitations. Newer Web search engines allow for intuition and complex algorithms to help people find what they meant rather than what they said. There is still, however, a good case to be made that the basics taught to searchers back when online began are transferable skills and should not be neglected in the Age of Web.

Letters to the Editor should be sent via email to marydee@infotoday.com.