Beyond
Google
Google is
seen by many as a search stanÂdard. Initially it built on its simplicity, huge
reach, and speed of response. Lately it is moving in many different
directions-from a directory-based search, to Google Earth to searching your
desktop. And, of course, adverÂtising based revenue streams. But if one were to
go back to the original proposition of search engines-finding of relevant
inforÂmation-then is Google really the best search engine? Are there more
around that offer, if not better, at least equally good search?
In general, any search engine
uses three basic components to answer a query: Its Web crawler crawls the Web or
visits Web pages regularly to gather information about them; the indexer
generates an index of Web pages based on the information gathered by the Web
crawler; and, finally, the query server displays your results in a particular
order.
Google uses proprietary
software, PageÂRank-developed by Google's founders, Larry Page and Sergey
Brin-which use text-matching techniques to find pages that are both important
and relevant to your search. The popularity-ranking method works out the order
in which the search results are disÂplayed-based on how many other sites link
to each page. MSN and Yahoo (and AltaVista) use categories and keywords to index
pages. The MSN index generator ranks those sites higher that are updated
regularly, have high volume content and have many sites link to it.
Ask Jeeves flagship search
sites, ask.com and teoma.com, work on the Teoma search technology. Based on the
subject of search query, Teoma organizes sites into naturally occurring
communities. Its Subject-Specific Popularity technique ranks sites by the number
of same-subject pages that are referenced by each site. Vivisimo's clusters
free the user from the burden of gleaning right information from the reams of it
available on the Web. This search engine gives high importance to the quality of
resulting cluster descriptions visible to the user.
The difference in searches
arises from the frequency and accuracy of the Web crawl, size, and type of
index, ranking methods and relevancy of information. Google is rated as the
biggest search engine with an index of more than 8 bn pages, as per the figures
available with the Search Engine Watch (an independent analyst of the search
engine industry). The second in line is MSN with 5 bn pages, followed by Yahoo
with an estimated 4.2 bn and Ask Jeeves at 2.5 bn. But, obviously, there is more
to search than size. Relevance matters.
And, on that parameter, there
is a long way to go. I tried some searches-'Natwar Singh Oil', 'Gandhi
Information Technology' and 'search engine industry size'-the first one
to see if the web put out any historical coverage of this issue. The second to
see if anyone had at anytime linked the Mahatma's economic philosophies with
information technology (a very long shot) and the last one to check if one could
get a quick, specific, answer on what is the size of this new industry (a very
optimistic shot). The search was tried on Google, MSN, Vivisimo, and Teoma.
The results were grey on all.
'Natwar Singh Oil' refused to give any more information than the current
news coverage from across the globe. 'Gandhi Information Technology' gave
many names of educational institutes, and the search engine query gave nothing
with pinpoint accuracy. There were many other advanced options that were not
tried. And this is not meant to be a comprehensive testing of search engines.
But one thing was obvious.
Search engines have a long way to go. So do the searchers. We all like quick and
clean information. That is simply not there. So the next time you are searching
for options, try the advanced searches and be imaginative. Who knows what you
may uncover?