SV: [SIGMETRICS] comparing web links to citation analysis is fall acious http://www .miislita.com/searchito/business-scene.html

[Nicolaisen Jeppe]

-----Oprindelig meddelelse-----
Fra: Garfield, Eugene [mailto:garfield at CODEX.CIS.UPENN.EDU]
Sendt: 4. august 2004 18:35
Til: SIGMETRICS at listserv.utk.edu
Emne: [SIGMETRICS] comparing web links to citation analysis is
fallacious http://www .miislita.com/searchito/business-scene.html


http://www.miislita.com/searchito/business-scene.html
----------------------------------------------------------------------------
----
05-13-GOOGLE- #1 Google Employee: Is PageRank Dead? Ask Craig Silverstein
and he will dismiss the whole issue with a royal dumb statement
a-la-Donald-Rumsfeld "The point of view that PageRank is dead is kind of a
very static view of the world." Oh, really? Bah! The fact is that (a) the
Web is a dynamical system, (b) the PageRank metric was full of fallacies
from the start ("users never click back", "this metric is nearly impossible
to deceive", "a link is a vote of importance", "users behaviors can be
modeled as a random walker", "the probability a user visits a page is its
PageRank", etc), and (c) SEOs and bloggers have proved that their metric is
easy to deceive. As we have mentioned for the last two years, PageRank has
been the biggest theoretical blunder from Stanford DB. With regards to links
as votes, read Mi Islita.com's 05-07 post. The common analogy in the sense
that "a link is a vote of importance" similar to literature citation, is
just a fallacy. Equating link popularity or links in general to citation
analysis is preposterous and a shame. Anyone familiar with Dr Eugene
Garfield's Citation Analysis Theory knows that citation analysis is driven
by peer-review examination, editorial policies and topic-specific
categorized publications. On the Web, where anyone can say almost anything
at any time, or add, remove and update links at will (or do keyword spamming
in links, pay for link advertisement or trade links) peer review is almost
absent. The shear volume of links on the Web is driven by commercial
interests. Period. The marketing analogy used by Google and SEO/SEM firms
with vested interests in the sense that "a link is a vote of importance" is
just that, a mere marketing line with no mathematical basis. The fact that
unethical bloggers and others have been able to "bomb" Google and unethical
SEOs are still gaming Google confirms the obvious. Yes. We do feel sorry for
Stanford DB theoretical patch makers and the Craigs of the World.

----------------------------------------------------------------------------
----
05-07-GOOGLE - IT Researchers: "PageRank is a Broken Model" Two years after
Mi Islita.com pointed out the many fallacies of the PageRank metric, here
and also here, IT researchers are now agreeing in one thing: PageRank is a
broken model. Back in 2002, we pointed out the obvious but many "seo/sem"
experts from UK to USA to elsewhere preferred to attack our grammar (or
"grammer"?) and play with their now useless "pagerank calculators". The very
same "experts" are now crying foul. Duh! As for the common analogy in the
sense that "a link is a vote of citation importance" similar to literature
citation, this is also a fallacy. Equating link popularity or links in
general to citation analysis is preposterous and a shame. Anyone familiar
with Dr Eugene Garfield's Citation Analysis Theory and ICI knows that
citation analysis is driven by peer-review examination, editorial policies
and topic-specific categorized publications. On the web, where anyone can
say almost anything at any time, or add, remove and update links at will (or
do keyword spamming in links, pay for link advertisement or trade links)
peer review is almost absent;links are mostly driven by commercial
interests. Thus the marketing analogy used by SEO/SEM firms with vested
interests in the sense that "a link is a vote of importance" is just that, a
mere marketing line with no mathematical basis. The fact that unethical
bloggers and others have been able to "bomb" Google and unethical SEOs are
still gaming Google confirms the obvious. We feel sorry for Stanford DB
theoretical patch makers. R.I.P. (Note. This post, updated on 05-09)





______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email
______________________________________________________________________