[Sigia-l] Search and classification

[Karl Fast]

> I don't fully agree with Jared that the end of Search is nigh,
> except in very specialized circumstances. I think search will
> remain a bedrock feature of the Web for a long time. However, I
> think that Jared makes some really important points about search.

Is the Web a specialized circumstance? I would argue that it is and
so something like Google is a special case. Why do I say this?

 - The Web is a huge freaking document set

 - There is no formal coordination between information providers
   (web sites) and search providers (eg: google, yahoo). This may
   change in some respects through things like metadata harvesting
   protocols, but not that much (shameless plug: I'm speaking on
   metadata harvesting at the IA Summit)

 - There is no way we will come up with a universally agreed system
   for tagging and classifying content, and even if we could agree
   to one, it would be too time-consuming and expensive to tag the
   Web based on this mythical standard.

In sum, the Web is big and messy, we can't change that, and we
wouldn't want to: this is a feature, not a bug.

HOWEVER, I think we have missed a key point in Jared's argument.

We should distinguish between "Web search" and "web site search."
The three factors I list can all be overcome (to varying degrees)
within an individual site, but not on the Web as a whole.

Our debate over Jared's research findings on search is almost
entirely related to the issue of web site search. We are muddying
the waters badly when we make a point and back it up by pointing to
Amazon, and then back up a related point using Google.

They are not the same thing at all.


--karl