At 11:06 AM 9/17/2003, Seth Dillingham wrote:
>Now I have a question: does anybody actually care about this stuff?
>Google's good, but does anybody actually expect it to be perfect, and to
>understand everything it reads? We know the rules that Google uses, so if
>it's important that you be ranked highly for some search term, then why
>don't people just follow those rules instead of accusing Google of cheating?
A lot of the Google backlash seems to be a backlash for the sake of having
a backlash. Google is more than just good -- it is amazing, especially if
people take the time to learn how to use more of its advanced search
capabilities.
The "problem" that Winer and others have is that there is *no* good method
other than human coding, to my knowledge, to return perfect results to
single term searches on general words. The problem isn't that Google
returns its own site for a search on weblogs, but rather that searching
just on "weblogs" and expecting to find a given sort of sit is sort of
ridiculous. What exactly is someone who searches just on "weblogs" looking
for? A list of weblogs? Information about starting a weblog? Background on
the sociological phenom. of weblogs? Criticism of weblogs? This reminds me
of an associate who was frustrated because he wanted didn't get high
quality results from searching on "rock" in Google (the same sort people
who go to animalrights.net and type "animal rights" in the search box).
Until a couple years ago I had assumed that the existence of things like
Google would increase people's knowledge of proper searching techniques.
Trying to find things in a physical library is often difficult and I can
understand why people found it frustrating, but on the Net you've got these
huge indexed databases and instructions on using them are just a click or
two away. But people seem to prefer to type in generic strings and hit the
"Feeling Lucky" option.
I'm not a big fan of Internet-izing education, but if there is one area
where schools could really make a big difference, incorporating a lesson on
how to get the most out of search engines into the standard research paper
class would be a very good thing.
Boolean -- it's not just for logic classes anymore!
