ISSUE 387: THE HIGH-TECH OFFICE--Alan
Zisman
Finding information on the World Wide Web
starts with locating the right search engine Mar 25
1997
Despite my occasional cynicism, I have to admit
that the Internet has its uses. But it can also seem like the world's
largest and most accessible vanity press (you know, those businesses
that charge wannabe authors to see their works in print).
And like seemingly millions of others, I, too, have
succumbed to the temptation to have a personal Web page -- it's a
chance to present my own view of my self-importance to the world.
Quickly now, go to stargate.vsb.bc.ca/total_ed/az.htm and
you'll see a photo of me, complete with glasses (unlike the picture in
the corner of this column), as well as a shot of my dog, Koko. Perhaps
of more value, there's also a page with links to more of my immortal
prose from this and other publications as it shows up in various sites
across Cyberspace.
Keeping that page current poses a problem, however.
The Internet is a fluid kind of place: new addresses appear, while
others shut down or change; new information gets posted without anyone
necessarily being informed.
A couple of years ago, I had the sense that the
Internet was like 50,000 libraries all hooked together, after someone
had destroyed the card catalogues. Things have improved, however. Since
then, a number of Internet "search engines" have appeared. These are
sites that try to catalogue the contents of the Internet, allowing
users to quickly get lists of addresses matching some criteria.
But it's easy for simple searches to get out of hand.
I recently went to search site Alta Vista (www.altavista.digital.com),
and typed my name into the search field. I'm always amazed at the speed
with which this search engine can prepare a report from 50 million or
more Internet sites. Within seconds, it told me that it found 831
matching "Zisman" (not a very common name, but there's a Michael
Zisman, not a relation as far as I know, who's a big shot at IBM),
and 485,588 matching "Alan." Putting them together, Alta Vista was
prepared to show me 4,000 Internet sites that might match my query --
10 at a time.
Most of the top 10 weren't articles that I'd written.
The big problem with search engines is learning to separate out the
junk. Luckily, there are ways to improve your search results.
Unfortunately, while all search sites offer to make your queries more
specific, each does it in its own way.
For example, on Alta Vista, if I type "Alan+Zisman" in
the search field, I only get responses that contain both words -- in
this case, narrowing the response from 4,000 to a mere 13. But when I
tried the same trick at the Lycos search site (www.lycos.com), I
got 21,160 responses -- without any of the first 10 actually referring
to what I was looking for (though one reference to Michael Z. showed up
amongst a bunch of other Alans).
At the bottom of the page, however, was an option to
search for either word or for both words. Choosing the latter option
got me exactly zero responses (known as "hits" in Webspeak) -- a
humbling experience.
A third search site, Web Crawler (www. webcrawler.
com/), is clever enough to give a hint right under the search entry
field -- type "Alan and Zisman" they suggest, or "Newt not Gingrich."
This got me two hits, but one of those turned out to be a page that
mentions some other Alan, and the oft-appearing Mike Z. in the same
article -- enough to get included in my search.
Perhaps the clearest of the bunch, for my purposes,
was HotBot (www.hotbot.com). Here, a drop-down list makes it
easy to clarify what you want, such as all of the words, any of the
words, the exact phrase, and more. Searching for "the exact phrase"
brought back 24 hits, all with my actual name in them, and most of them
linking to things I had written that had been posted somewhere on the
Web.
Here are a couple of lessons that may be of value even
when searching for something other than my collected works:
* While Web search sites are attempting to organize
and categorize the vast amount of data available on the Internet, no
single site has it all. You may have to check several to find what
you're looking for.
* Simple queries often turn up what you want, mixed in
with literally thousands of unwanted pieces of data. Learn a few tricks
to help narrow down your searches. Unfortunately, each search site does
this differently.
* If I had to pick a single search site, my current
favourite would be HotBot. This site combines a large amount of
information with the most helpful interface, making it the best of the
current crop at getting just the data you're looking for.*
|