Alta Vista coverage much smaller than thought?

[Robert C. Williford]

Prentiss Riddle wrote:
> 
> I'm curious what readers of Web4Lib think of John Pike's discovery that
> Alta Vista retains only a small sample of web pages from large web
> sites.  He says that Alta Vista has only about 600 pages from his
> 6000-page site, and admits to having only 300 pages from the
> 300000-member Geocities ISP!
> 
This was no surprise to me, since three students here at the University
of Missouri-Columbia School of Library and Informational Science (since
graduated) wrote a book based on their in-depth analysis of search
engines.  They gave a presentation to us about 3 months ago--maybe
longer.  One of their conclusions was that web designers should make
their sites wide and shallow (in terms of hierarchy, not content!),
since some search engines only look at the top few levels--believe four
deep was about the max.  This is counter-intuitive to good graphic
design, where you might want to have a nice intro screen (first level),
which links to a table of contents (level 2), which links to an intro
page about some info item (level 3), which then offers another page of
links to real information (level 4).  By the time you get to the gems of
information, you're deeper than the search engines' knowbots ever dug. 
Probably didn't get "catalogued."  

I only recall the name of one author, David E. Moxley, who's at
David_E._Moxley at muccmail.missouri.edu.  Believe that the title of the
book is "Neal Schuman Authoritative Web Searching."  I know the ISBN is
supposed to be 1-55570-307-0.  It should be out in late May, but I think
Neal Schuman is already taking orders.

Obviously, I haven't seen the book, but the presentation I saw had lots
of solid info that was really eye-opening.

Bob Williford
SLIS Student
University of Missouri-Columbia