Constrain Web searches to free yourself from muddling through muck
I know, it's not very catchy. But it works. Here's how. Constraining can help tweak the way you use Internet search engines, giving you a dozen good hits instead of 2,000 pointers to documents that vaguely resemble your topic. Directing the search engine toward words contained in a particular Hypertext Markup Language tag is an underutilized constraining tool. In a normal search, only the regular text of the document, not the text inside HTML tags, is
I know, it's not very catchy. But it works. Here's how.
Constraining can help tweak the way you use Internet search engines, giving you a dozen
good hits instead of 2,000 pointers to documents that vaguely resemble your topic.
Directing the search engine toward words contained in a particular Hypertext Markup
Language tag is an underutilized constraining tool. In a normal search, only the regular
text of the document, not the text inside HTML tags, is searched.
Let's say I want to find information on Internet hackers, but I want only information
collected at military sites. Simply entering ".mil " and "hackers"
returns thousands of documents.
But constraining the search with "domain:mil and hack*" turns up about 900
documents that contain the terms "hack," "hacker" or
"hacking." You can refine your search further by adding subject specific terms:
"domain:mil and hack*" and "subject:sendmail."
You can constrain searches via almost any common HTML tag. Looking for the Agriculture
Department logo? Enter "image:usda." Curious about who may have links to your
site? Enter "link:yoursitename.gov."
Visit http://altavista.digital.com/cgi-bin/query?pg=h
for a long list of ways to constrain searches. I get better results from constraining than
I do from using the Java topic map that Alta Vista offers.
The best way to make sure you find the nuggets of gold in the slag of the Internet is
to find a consultant, an expert who will prefilter information for you. Some of the best
experts on the Net are volunteers who maintain the frequently-asked-questions (FAQs)
lists.
How to find the FAQ folks? Start with a visit to http://www.cis.ohio-state.edu/hypertext/faq/usenet/FAQ-List.html.
The site has a list of known FAQ sites for UseNet newsgroups, complete with details on who
maintains the lists.
Many of these people maintain Web pages that link to news sources on particular
subjects. Isn't it great when others do the legwork?
Another good way to find information is to join or build an Internet community that
will keep you informed about your areas of interest. Communing usually requires that you
combine mail distribution lists, UseNet newsgroups and Web pages that serve your
particular needs.
How do you build such a community? Start by visiting http://www.lsoft.com/lists/listref.html,
where L-Soft International Inc. of Landover, Md., maintains a list of mail lists. L-Soft
develops and markets the popular Listserv software used by many mail list managers.
You can download a trial version of the Listserv software or browse through thousands
of lists to find a community that interests you. Some have several thousand subscribers.
You might also want to monitor the UseNet news group news.answers to look for new
and updated FAQ postings.
With Internet push technology, the basic idea is that the information you want comes to
you as it's available. You don't have to go looking for it.
Depending on whom you talk to, push technology is either a revolution that will make
Web browsers obsolete or a bandwidth pig that's attractive mainly to advertisers.
Push is not e-mail. It's an interactive window that supports Web protocols. Push
technology has been around for a few years on Web pages that update occasionally as long
as you are viewing that page. Recently, push services are showing up as Java or ActiveX
windows that monitor a news feed without upsetting other desktop work.
For a look at an elegant push solution, visit http://www.nando.net/nt/newswatcher/
to see the newswatcher Java applet designed by the Nando Times. Compare Nando with
PointCast at http://www.pointcast.com. I
find PointCast bandwidth heavy and intrusive compared to Nando.
BackWeb Technologies of San Jose, Calif., at http://www.backweb.com,
provides a series of push-updated public channels that you can monitor for news and also
has server technology that lets you set up your own push server and become a channel
provider.
The Castanet transmitter and the client-side Castanet tuner are available from Marimba
Inc. of Palo Alto, Calif.
Any the four approaches-constrain, consult, commune and push-will help keep you
informed. Together, they will shave many minutes from your data gathering, which is a lot
in these times of information overload.
Shawn P. McCarthy is a computer journalist, webmaster and Internet programmer for
GCN's parent, Cahners Publishing Co. E-mail him at smccarthy@cahners.com.
NEXT STORY: Health records project hits pay dirt