P1: IML
Wisman WL040/Bidgoli-Vol III-Ch-59 August 14, 2003 18:3 Char Count= 0
HOWTOSEARCH—THESEARCHER’SVIEW 725Figure 1: An overview of a search engine and its relation to Web sites and information
searchers. The basic steps to Web search are as follows: a search engine spider visits a Web
site to retrieve and index pages for later searches, the searcher sends queries to the search
engine and receives a list of page locations, the searcher selects and retrieves pages from the
Web site.as one of the most important means for finding informa-
tion (Nielsen, 1997).More than half of all Web site visi-
tors navigate predominately by search and use search as
their first choice to find specific information. Many us-
ability experts argue that search is so important it should
be available from every single page on a Web site. Part of
the rationale is that once visitors navigating connection
links get lost they will need search and should not have to
search for a search page.Types of Search
Search often fails to some degree. To identify good
search strategies it is worthwhile to consider the differ-
ent types of search and how search engines match differ-
ent strategies. Information retrieval experts (Rosenfield &
Morville, 1998) recognize four ways in which people
search:Known-Item Search:A search engine is not always nec-
essary. Searching a brokerage site for the latest stock
quote is one example. The information sought is uni-
que, is easily recognized when found, and often exists
at known Web sites. Searching, if any, is short and di-
rect and the answer immediately recognized. Search
is unneeded if the information location remains fixed,
even as the information changes. The search can be
restricted to specific sites, reducing search time and
improving the information quality.
Existence Search:Consider the difficulty in searching
for something that may not exist. Suppose that some-
one who needed to organize phone number lists, notes,
and appointments had never heard of a personal digi-
tal assistant (PDA). The person would know the needed
function but not where to look or that a PDA even ex-
isted. Search can meander as users discover and di-
gest information; even recognizing the answer when
found may be difficult. Web search engines are a natu-
ral choice for determining whether information exists
because they attempt to find all available information
on the Web. Entering a few well-chosen query words
to a search engine will search millions of pages andcan produce thousands of pieces of information. One
gap in existence searching is that search engines cover
only a fraction of the Web; the more obscure informa-
tion can exist on parts of the Web never visited by the
search engine.
Exploratory Search:This is the case when the user knows
that the subject such as a PDA exists but does not know
exactly what a PDA can do and need to learn more. Un-
covering and comprehending the information may be
long and circuitous; recognizing the answer may not be
immediately possible due to the lack of subject knowl-
edge. The most productive search strategy is brows-
ing information organized into subject lists with main
subject areas and a breakdown into finer categories of
related subjects.
Comprehensive Search:This type of search is best suited
when users know the search subject, can recognize the
answer when found, and need all available information
on the subject. For someone with a general knowledge
of PDAs, buying a PDA requires comprehensive infor-
mation on several brands, models, and suppliers to
make an informed decision. Web search engines work
well for comprehensive searching of the entire Web but
for broad subjects such as PDAs can produce an over-
whelming volume of information.Information Sources
Most successful searches fail to some degree at the first at-
tempt. Other than the known-item search, most searches
evolve to use several search strategies as one attempts to
iteratively improve the quality of information retrieved.
Because Web search sites accommodate one or more of
the search strategies we will focus on the type of sites best
suited to users’ search needs, the most common ways to
search, and techniques for successful searches.
Though the term search engine often applies to any
computerized information source, there are at least three
different types of Web information sources: automated
search engines, human-organized lists, and portals. Most
information sites are a mixture of some or all three types;
the following characterizes each.