[Sigia-l] metadata redux

[Karl Fast]

> > These core design assumptions are *completely at odds* with the
> > realities of the distributed information environment found on the
> > World Wide Web today.
> 
> Nooo. The web actually realized these anomalies early on and went
> around them.

Ahh, but you see that is Lynch's point. He agrees with you.

He is pointing out to the academic IR and LIS community that there
are some traditional assumptions made in classic IR that do not
apply in the web environment. These assumptions, he argues, are so
deeply ingrained that we don't even think of them as assumptions.

This is why web search engines relied on HTML metatags in the first
place. It took a few years before vendors realized they were
unreliable. They came from the classic IR world were the assumptions
Lynch is talking about are so deep they're not even viewed as
assumptions. How could it be otherwise? The web is how.

Now that's for things like Google and altavista and hotbot.

As you note, things like Epinions and Amazons user reviews and Ebay
seller ratings are also reliant on user supplied metadata. And in
those systems it works.

Those mechanisms are different, I think. Gotta run, late for a
meeting, but brainstorm over why they are different.

--karl