[Sigia-l] Disposable navigation systems

[John O'Donovan-INTERNET]

Ted Han:

> So, ultimately your ability to allow for, or facilitate radically 
> different types of navigation/transformation over the same data set is 
> going to depend entirely on whats -in- the data & metadata, not some 
> layer of configurability sitting on top of your content.

On a related issue, thought people might be interested in this...courtesy of EPS...

TOP 4 SEARCH ENGINES ARE MORE DIFFERENT THAN YOU'D THINK

* A new study has shown that the overlap in terms of first page search 
results for the same queries between the big four search engines is fairly 
small.  What implications does this have for current search engine 
marketing and for future developments in this sector?

by Nick Dempsey, Analyst

According to the meta search engine Dogpile and researchers at the 
University of Pittsburgh and Penn State University, the footprints of the 
four major search engines are quite different - and are becoming more so. 
In a July 2005 study, which follows on from a similar one in April 2005, 
the researchers considered the first page of results for 12,500 random 
queries.  The first page was chosen because Dogpile's owner, Infospace, 
calculated from log file analysis that 89.8% of user search activity 
happens on that first page of results.  The likelihood that the same 
result for a given query would appear on the first page of results for all 
four engines (Google, Yahoo!, MSN and Ask Jeeves) was found to be only 
1.1%.  And the results comparing groups of two and three search engines 
are perhaps even more surprising: only 2.6% of results appeared on the 
first page of three of these engines and 11.4% were shared by two. 
Furthermore, a high proportion of total first page results for the same 
queries were unique to the search engine concerned: 73.9% of Ask Jeeves's 
first page results were unique to it, 71.2% of Yahoo!'s, 70.8% of MSN's 
and 66.4% of Google's.  The study refers to four different "editorial 
voices", created by different ranking and crawling software, where most 
users believe that they are getting a broadly similar experience from all 
the search players.  The overlap between these engines is even less than 
in the April 2005 study, where only Google, Yahoo! and Ask Jeeves were 
considered.

One possible explanation for the growing difference (unexplored in the 
Dogpile study, which focuses on the consequent benefits of metasearch) is 
the proportion of the measured web that search engines cover.  Antonio 
Gulli of Università di Pisa (also the director of advanced products for 
Ask Jeeves) and Alessio Signorini of the University of Iowa, estimated 
that in  January 2005 that there were 11.5 billion pages in the surface 
web, and of those, according to the study, Google indexes just under 70%, 
Yahoo! covers 57%, Ask Jeeves 46% and MSN 44%.  Clearly there is plenty of 
"room" in the web for the different engines to return different results.

If recognised, the increasing difference in coverage suggests that more 
users are likely to want to search more than one search engine when 
pursuing a piece of information.  It is not necessarily in the interests 
of the search players to highlight the relatively small levels of overlap 
between each other, since they have built loyal user bases, particularly 
Google, who remain convinced that they are getting everything important on 
the web from their one favourite destination.  But if users do become more 
likely to have multiple search toolbars or use metasearch, this should 
focus the attention of search engine marketers on spreading their ad 
budgets across the engines and optimising their sites for all of them.  It 
may also be that if users become more accustomed to using several search 
engines, they will also be more likely to consider the high-quality 
information databases such as Factiva, LexisNexis, Ovid or Gale, that 
might be available to them in a professional or academic setting.

The image of advertising-driven information companies with distinctive 
voices, through which users construct their information picture of the 
world is reminiscent of the rise of print newspapers, even though the 
"voices" of search engines are broadly accidental.  This image is 
complementary to Matt Thompson and Robin Sloan's EPIC 2014 vision, where a 
constantly evolving mass of data, categorised democratically and crawled 
by software which can pull out coherent stories from multiple sources, 
destroys the newspaper industry.  There may be several different flavours 
of "Googlezon".

© Electronic Publishing Services 2005

Related links
-------------------------------------
Dogpile-sponsored overlap study :: 
http://comparesearchengines.dogpile.com/OverlapAnalysis.pdf
Gulli and Signorini web size study :: 
http://www.cs.uiowa.edu/~asignori/web-size/
Searchenginewatch article covering Dogpile study :: 
http://searchenginewatch.com/searchday/article.php
EPIC 2014 :: http://www.robinsloan.com/epic/

-------------------------------------

 ::: John O'Donovan
 ::: Principal Technologist, Factual and Learning
 ::: Room 3306, BBC White City
 ::: 201 Wood Lane, London W12 7TS
 ::: +44 (0) 780 313 6620
 ::: john.odonovan at bbc.co.uk
 ::: www.bbc.co.uk


http://www.bbc.co.uk/

This e-mail (and any attachments) is confidential and may contain
personal views which are not the views of the BBC unless specifically
stated.
If you have received it in error, please delete it from your system. 
Do not use, copy or disclose the information in any way nor act in
reliance on it and notify the sender immediately. Please note that the
BBC monitors e-mails sent or received. 
Further communication will signify your consent to this.