Computers
may be powerful but they are not very smart. They cannot understand
human language. You cannot talk to a computer and ask it to find
the information on the types of deciduous trees native to San
Francisco. The computer does not speak English. The search
engine does not speak English.
Understanding
how to perform sophisticated searches for online information will
greatly increase your chances of finding what you want. Most search
engines let you define your search criteria in very specific ways,
but not all function the same way. The following are some simple
and common ways of defining your search criteria:
Keywords
A keyword or keywords are the words that the engine will use to
find pages. Generally speaking, the more specific the word or
words are, the better your returned results will be, because the
page that contains that word is more likely have something of
interest to you. Example:
using the term beagle rather
than dog is more likely to
return only pages with beagle information.
Using
more than one keyword can also help the engine narrow down pages
for you BUT if you are going to use three or more words it is
probably best to use Boolean logic (see Advanced
Search Techniques) and an advanced engine that will understand
Boolean logic, otherwise you will not be able to control how the
engine interprets the relationship between the words and how it
applies its search logic and ranking logic to them (see "phrases"
below)
Phrases
When you have two keywords that should be searched for so that
they are found together and in a specific order, like batting
average or financial aid
it is best to control the search engine so that it knows these
words need to be together in its searches. Some search engines
allow you to do this by choosing a button or a pull down menu
to select a "phrase" search or typing the words in a
"phrase" box. Advanced search engines usually let you
enclose the words inside quotation marks to designate them as
a phrase.
If
you do not control how the engine interprets the two keywords,
you never know if it sees the words as separate words (which may
or may not be in the same sentence) or as a phrase. Its search
logic may return results based on, how many times the first word
is found, how many times the second word is found, how many times
both words are found, how many times both words are found located
together ….or some combination of the above.
Example: Say
you are looking for information on mountain
biking. If you do not control the engine to ensure
that it searches for these words together you might get some webpages
with only the first word, some with only the second, some with
both words but not together (in separate sentences or even paragraphs)
and some with the words together in a phrase. Not only that,
but how it ranks the pages might also be problematic – one page
with only the word “biking”, but with it 321 times, might be ranked
towards the top.
Capital Sensitivity
If a search keyword is capitalized and the seach engine
is capital sensitive (not all are), it will return only documents
containing the capitalized word. For example, if you were interested
in documents relating to the country of China,
capitalizing the word and using an engine which supports capital
sensitivity would narrow down the number of results returned,
eliminating documents which relate to china dishes or cookery.
In
many instances however, it is better to leave keywords uncapitalized
to allow the engine to return results of documents which have
keywords in either form.
Abbreviation
or truncation for root words
If you are looking for a keyword that might have many variations
based on a root word and you use an engine that allows abbreviation
or truncation – you can truncate or abbreviate that word down
to its root to enable the engine to search for all possible variations.
Example:
Say you are looking for information on gardening and so you use
the word “gardening” and get
25 results – none are what you need, though. Then you use the
word “gardener” and get 13
results – none are what you need, though. Then you use the word
“garden” and get 17 results
– none are what you need, though. Then you use the word “gardens”
and get 15 results......and finally find pages you want.
An
easier and quicker way to execute a search for all or these versions
of a root word (garden) would have
been to use an engine that allows abbreviation and employ the
symbol * to truncate the word down just far enough to include
all versions of the root word that you may want – garden*.
NOTE:
you don’t want to abbreviate down too far, as in this case, if
you truncated down to gar*
you would get literal and figurative garbage…and
gargoyle and garage
and gargle.
"+"
require and "-" exclude
Some engines offer a variation of the Boolean operators AND and
NOT (see Advanced Search Techniques).
A "+" symbol preceding a word (with no space in between)
will require that the word be present in documents. A "-"
symbol preceding a keyword will ensure that the word is not present
in returned documents.
NOTE:
All words which must be in the document should be preceded by
a "+" symbol, even the first word. Example: +resistance
+antibiotics ensures that both resistance
and
antibiotics are in returned results.
|