USING SEARCH ENGINES AND DIRECTORIES TO FIND INFORMATION ON THE WEB
Once you are online and able to browse the web, the next step in
research is to use Web search tools to locate
relevant information. This section will look at the steps and resources
available for searching.
Devising a Search Strategy: Step 1: Describe Your Topic
Before you begin keyword searching, it can
be profitable to spend some time thinking carefully about your topic, brainstorming
keywords and other terms you might use in your search. At this stage you
should think creatively and broadly of synonyms, spelling variations, and
key people or organizations associated with your topic. For example, if
your topic is "physician assisted suicide," possible keywords might include
"euthanasia," "mercy killing," "right to die movement," "jack kevorkian,"
and "hemlock society." You might also think about common occurrences of
your keywords that are not in the context you intend. For example, you
might want to try to exclude search results that relate to euthanizing
animals, another common occurrence of the term "euthanasia." This initial
thought about your search can help you get started at a search tool, and
will ensure that you have variations ready if your search is not immediately
successful. This list is just a starting place however; once you begin
searching, you will undoubtedly find other variations to try.
Devising a Search Strategy: Step 2: Selecting an Appropriate Search Tool
What Are Search Tools?
A search tool is a site on the web that is dedicated to helping you find
other sites. When you search at a search tool, you are not searching the
web directly, but rather the database of the sites that particular tool
has compiled. There are several different types of search tools, and we
will look at examples of each, but first let's consider some generalities
that apply to all search tools:
-
No search tool includes the entire web, and overlap
between what the search tools include may be limited.
-
Each search tool will support a different syntax, or set of rules for searching,
with varying degrees of sophistication and flexibility supported.
From the above points you can conclude that to be as comprehensive as possible
you will need to search with more than one search tool. You may also conclude
that given that the rules for searching will vary from tool to tool, it
is probably most efficient to spend significant time learning the syntax
and becoming efficient at just a couple or three tools. For most searchers
it is simply too daunting a task to become an expert searcher at all search
tools, while being able to use a couple of tools well will yield good results.
Of course, if in any given search you have trouble finding information
on the tools you know best, you can extend your search to other tools.
Tip: To help you stay current on what the different search tools
are and how they compare to one another, visit the site Search
Engine Watch.
Search tools may be described broadly by whether they are one of two
types: directories or search engines.
Directories
Directories, also called catalogs, may be characterized as relatively small
databases, compiled by humans rather than machines, with sites classified
and organized into subject groupings. An example of a recommended directory
is Yahoo, one of the first and best.
Other good directories include: Librarians
Index to the Internet, and about.com. Directories have a "top down" menu structure,
moving from broad subjects through more specific subjects and eventually
to specific sites. You can therefore browse through directories,
moving through increasingly more focused topic areas to get to sites on
your subject. For example if your ultimate aim is to locate sites about
Internet Addiction, you might start on Yahoo's Directory
by clicking on the broad topic of "Health," then selecting "Diseases and
conditions," then "Mental Health," then "Addiction and Recovery" and finally
"Internet Addiction." Under this final subject heading you would find links
to web sites related to this topic.
"Tunneling down" through subject terms in this fashion might be a good
approach if you are not sure what you are looking for, but it is usually
faster to approach a directory by searching it by keyword. For example,
if you search on the keywords "internet addiction" at Yahoo, the search
results will include first a list of categories (subject headings) that
include these terms, then a list of sites that match the keywords. Clicking
on any one category will give you a listing of all the sites the Yahoo
catalogers have determined relate to that subject.
The advantage to using a tool that employs subject classification is
that all the sites listed under a particular category have been selected
by a human as really being about that subject. This human intervention
stands between you and the kind of odd search results you get at regular
keyword search sites. Because directories tend to be smaller and more selective
in the sites they include, the quality of the sites may be better than
what you would find in unmediated searching. Directory sites like Yahoo
are also a faster way to search for a specific company page than the larger
search engines are: use them when you don't want to find every random mention
of a particular company, but want a good chance of finding the company
home page itself. In other words, directory sites are good for "telephone
book" type searches - finding the web page for a particular company or
organization. They are also a good place to start any research, for finding
a neat, compact starting set of sites on a particular topic.
Practice: Go to Yahoo's Directory and find a category that deals with the death penalty. Look through a few of the Web sites in the category and decide if you might be able to use them in a research paper.
Search Engines
Search engines, in contrast to directories, may be characterized as much
larger databases that lack "human touch" subject classifications. Remember
that when you keyword search at a search engine you are not searching the
web directly, but rather the database of web sites they have compiled.
There are three parts of a search engine:
1) A "spider," also called a "crawler" or a "bot" - the program that
goes out foraging around on the web for sites, explores them, and brings
them back.
2) The huge index or database created by compiling the pages the spider
retrieves
3) A web interface/program to receive your search request, compare
it to the index, and return a set of results (links to sites) to you.
An example of a recommended search engine is Google.
Compared to the Yahoo directory, Google has a much larger database of over eight billion pages as of 11/22/04 (the
homepage of Google lists the current number of pages indexed in the search engine).
Google is compiled and maintained by machine rather than
by humans. Since the Google search engine is so large in scope, you
will want to be as specific as possible when you search it, using the narrowest
possible search terms for your topic. (Google does now also include
its own directory, a topically arranged subset of the sites available by
keyword searching, but the main focus of the tool is the search engine).
Syntax Notes for Searching Google (and Many Other Keyword Search Tools)
These syntax notes refer to Google's Main Search specifically, but most
of this syntax is fairly standardized and would be valid on many other
search tools. To be sure you are using the right syntax at any given tool,
you should always check the help section at the site.
-
The basic logic of a Google keyword search is that the engine will
retrieve documents with any of your keywords, but documents with all of
your keywords will display first in your results list. For example, if
you type the keywords:
addiction internet
You will retrieve documents with just "internet" or just "addiction"
as well documents with both of these terms - but documents with both terms
should display first.
-
Use quotation marks to search for a phrase. For example:
"internet addiction"
will bring back only documents where those two words appear together
in that order.
(Note: Google has recently introduced a "semi-automatic" phrase
search; if it recognizes your search terms as a phrase, it will automatically
search as a phrase, with or without quotes. To see this semi-automatic
phrase searching in action, compare what happens if you type 1) internet
addiction 2) "internet addiction" and 3) addiction internet)
-
You can require that a word be present in your documents by putting a +
in front of it. For example:
"internet addiction" +chat
would bring back only documents containing the word chat (the ones that
also contained the phrase "internet addiction" would be at the top of the
list).
+"internet addiction" +chat
would bring back only documents containing the word chat AND the phrase
"internet addiction."
internet online on-line +chat
would bring back documents that contain chat, and will also look for
the variations (internet, online, on-line) included
-
You can require that only documents that do not contain a particular word
be included in your results. Do this by putting a - in front of the term.
For example:
"internet addiction" -chat
should exclude documents that discuss chat. (Remember: computers are
very literal - this search would also exclude a document that began: "This
article will not focus on chat rooms.")
-
It's better to use lower case than upper case - upper case forces a case
match. For example:
internet
will match Internet and internet, while
Internet
will only match Internet.
Test Yourself
Practice: Go to Google and search for the phrase "death penalty" (remember to include the quotes). Is there a difference between the Web pages you find here, and the Web pages you found in the Yahoo category above?
Special Kinds of Search Engines: Meta-Search Engines and "Natural Language"
Meta-Search Engines
One special kind of search engine is called a "Meta-search engine." These
search tools allow you to submit your search to several search engines
at once. This can be useful if you are searching for a relatively obscure
piece of information, or to get an initial sense for how much information
is available on your topic. However, since your search terms are being
submitted to numerous engines at once, and since as we have learned each
search engine has different syntax rules,
your search at a meta-search engine must be simple enough to be understood
by all. If you want to be able to use any advanced search features, you
will be better off going directly to one or more single engines.
Recommended meta-search engines:
Ixquick
Dogpile
Practice: Go to Ixquick and search for pages that deal with the death penalty. Again, note any differences between Inference Find results, and the results from the other search tools you have used.
Natural Language Searching
Another kind of specialized search engine allows "natural language" searching.
In other words, you can enter your search with a natural English question,
and the engine will attempt to answer your question.
This kind of searching works wonderfully when it works, but at this
stage the technology often does not work as expected. When the engine understands
your question correctly, it can save you a great deal of time. However,
when the engine does not understand your question, you are no better off
than you would be at a regular keyword search. As an example, suppose that
you are trying to answer the question, "Why is New York city called the
Big Apple?" If you try to search for this answer at a keyword search site,
your keywords might be: "new york city" and "big apple." Image the number
of hits you would get for this search (or better yet, try it!). You'll
get numerous businesses located in New York with "big apple" in their names,
sites about being a tourist in the city - any site at all that happens
to include these terms, in any context. Try to included a term like history,
or "called" or some other way to focus in our your question, and you'll
see how difficult this can be. You might be able to answer your question
at one of the sites you've retrieved, but it will certainly take some digging.
Now try a natural language search at a tool that supports this technology,
AskJeeves
(just type Why is New York city called the Big Apple? Into the question
box) and see how nice it is to cut right to the chase. But, try a more
complex question, or one Ask Jeeves doesn't happen to know about, (What
are the privacy concerns surrounding supermarket club cards? Are cocker
spaniels overbred in the U.S.?) and you'll get less satisfying results.
Practice: Go to AskJeeves and ask; "Do you have any information about the death penalty?" See if you like the way AskJeeves presents information about a topic.
Previous | Page top | Next
Please contact
librarians@ohlone.edu
with your questions, comments, and suggestions.
Last updated 11/19/04. Terms and Conditions of Use
& Disclaimer.
Copyright © 2004 Ohlone College. All
rights reserved.
