[Sigia-l] ROI/Value of Search Engine Design

[Jim Kauffman]

I think "search" is being defined very narrowly here. In most cases (expert
researchers aside), the goal is to find minimally useful and relevant
information about what the searcher is interested in. Expectations are low,
and for good reason. We can't profile anonymous searchers today, so we need
to create a variety of tools optimized for the task, whatever the user's
goals might be.

Ben Shneiderman and his team at the University of Maryland have done
wonderful work imagining new ways of finding things. They rightly
concentrate on creating new tools for navigating huge bodies of data, based
on new ways of addressing the variety of searches that users want to do.
We've invented the wheel, but we certainly haven't imagined the automobile
yet, and it will probably take many more years of research and testing until
we can.

Perhaps at this point we lack the language to adequately discuss
"searching", much less to analyze it. The desire to find answers in data is
almost as complicated as the desire to find answers in our larger lives.
Google is not the "killer app" of searching, but it increases our
understanding of what people want from a simple search. Unfortunately for
most of us, the simple search is the only tool we can easily use today.

-Jim Kauffman