|
About ITP
Childhood ITP
Teenage ITP
Adult ITP
Medical Advisors
Research into ITP
Events
Convention
Seminar
Send for:-
Publications
American Perspective
Downloads
Leaflets and more....
Search this site


How we spend your donations
|
 |
Webwatch
June 2003
As included in the last issue of The Platelet, mention was made about
how to search for a word in just one site, using the syntax child site:www.itpsupport.org.uk
in Google.
Perhaps some more information on searching the internet may be of help.
There are two kinds of search engine, one that relies on computers to
index words, the other that relies on humans to make an index of websites.
The second kind is declining in use, they require lots of people to create
and maintain the index. Examples of these are Open Directory or LookSmart.
Of the first type, engines such as Google use computers to look into (or
spider) websites, making a gigantic index of the words it finds. It does
not "understand" about subjects. For example, a recent search
for ANA and ITP resulted in 7,500 hits, the first being a link to a site
that said "On behalf of IT Power, I send my sincere condolences to
Ana"! The software was just doing what it was told, it found "Ana"
and ITPower! Had I known that in the context of ITP, ANA stands for antinuclear
antibodies and had searched for "antinuclear antibodies itp",
the result would have been less than 200 largely relevant hits. The lesson
here is that a little knowledge of what you are searching for is very
useful. Successful searches come from using subject or language skills
than "computer knowledge".
Search engines vary all the time, currently the popularity of the top
ones are approximately Google with 30% of users, Yahoo 29%, MSN, 27%,
AoL 10%, with Lycos, Looksmart, Ask Jeeves etc. making up the balance.
The oldest engine, Yahoo, used humans to build their index until recently,
when it linked up with Google. Search engines also use each other's indexes
to find results and many use systems that do not have a public face such
as Inktomi. This is a huge index of websites only available to other search
engines. It means that means the majority of the people searching the
internet are actually searching essentially the same index of information.
Naturally, searching for "itp support" in Google, Yahoo or MSN,
the Association website comes out at the top!
Sometimes, the results of a search seems to give information totally
different from what was searched for. Travel companies and hotel reservation
companies seem to be the worst at this, but how does it happen? As search
engines simply look at words on the site, especially words near the top
of a page, unscrupulous people have been known to put all sorts of commonly
searched for words in the website but use the same text colour as the
background so you cannot see them! The search engines dutifully add these
words to their index. Sometimes a large area that appears blank actually
contain these words. Avoid such sites, they have an ulterior motive for
their web presence.
As an example of effective searching, you may like to set yourself the
task of finding out what substances in chocolate may be related to making
some people feel unwell. A search for "chocolate ingredients"
will reward you with over half a million recipes! A search for "chocolate
chemical ingredients" will give nearly 30,000 hits, too many to be
useful. Adding "migraine" to the search gave 1,300 hits which
is better but still too many. Reading some of the first sites from this
list however, gave a clue to what to search for, such as phenylethylamine.
Searching for "chocolate chemical ingredients migraine phenylethylamine"
gave about 70 very useful hits.
A very detailed account of how search engines work etc. can be found
on www.searchenginewatch.com
Happy surfing
Howard
Index to Webwatch
|