[LINK] The Real Google Inventors - Was - Google Moves to Take Back Android

[Tom Koltai]

> -----Original Message-----
> From: link-bounces at mailman.anu.edu.au 
> [mailto:link-bounces at mailman.anu.edu.au] On Behalf Of Ash Nallawalla
> Sent: Monday, 23 May 2011 8:36 PM
> To: 'link'
> Subject: Re: [LINK] Google Moves to Take Back Android
> 
> 
> > From: Tom Koltai
> > Many around the world believe that Larry Page invented the 
> Page rank 
> > reference model for Google search. Most also believe <snip>
> > 
> > Both of these groups would be wrong in their beliefs.
> 
> Who do you think invented PageRank?
> 
> Larry alone didn't invent PR but it is named after him. Some 
> say the lawyers recommended using his surname to get added 
> justification for protection.
> 
http://www.seobythesea.com/?p=5233
http://en.wikipedia.org/wiki/PageRank
http://www.google.com/corporate/tech.html
http://www4.ncsu.edu/~ipsen/ps/simax69812.pdf 
http://infolab.stanford.edu/~backrub/google.html
http://infolab.stanford.edu/~backrub/pageranksub.ps (PostScript format)
etc

All very impressive Ash... 

For most inventions in our history there have been the inventors, and
then the "popular inventors".

E.g.:
Tesla vs Marconi  (wireless)
Édouard-LéonScott de Martinville vs Thomas Edison (phonograph)

But let us for a moment look beyond the "popular" and discover the real
inventors of Page Rank...

For the History of Google's search engine, one has to go back to:
William Tunnicliffe's suggestion in 1967 to separate the information
content of documents from their format.
Followed by Goldfarb's implementation of GML in 1969 and before that to
Ted Nelson's ill-fated by highly prophetic Xanadu project.

The Google Page Ranking methodology was first identified by Cooper in
1973 and named the Probability Ranking Principle, PRP [and his later
work on "expected utility" (relevance) in 1978].
In Kowalski 1997, Kowalski examines the statistical approach of the
relevancy of information to the querant using probability based on the
number of databases a particular topic is referenced in.

Kowalski (and others) called into question some of Coopers conclusions,
but as this appeared to be based on the available computing capabilities
of the differing points in the Search Timeline, I have discounted this
as not being relevant.

What is curious is that PRP is not mentioned anywhere that I can find in
Page's published works although his algorithm is based on:

Quote/ [From: The Anatomy of a Large-Scale Hypertextual Web Search
Engine Sergey Brin and Lawrence Page]

PR(A)=(1-d)+d(PR(T1)/C(T1)+...+PR(Tn)/C(Tn))
Note that Pageranks form a probability distribution over web pages, so
the sum of all web pages' pageranks will be one.
/Quote

So the most important references are omitted from the research paper.
Probably an accidental oversight.

What is admitted is that PageRank itself was based on citation analysis,
developed initially by Eugene Garfield in the '50s at the U-Penn, and by
Hyper Search, developed by Massimo Marchiori at the University of Padua.

In 1996, Robin Li, created Rankdex (owned by IDD info services) and was
developing the ranking and scoring of internet pages.

Quote/ [From: http://www.rankdex.com/about.html]

About Rankdex

RankDex was established in 1996 by Robin Li who previously worked for
Infoseek, an Internet search engine in the early years, as an engineer.

RankDex was the leading Internet search provider from 1996-1999. RankDex
was the first search engine that used hyperlinks to measure the quality
of websites it was indexing. Many experts and commentators are unaware
that RankDex was cited in Lawrence Page’s patent application for Google
PageRank as the first qualitative search engine. In 1998 RankDex was
acquired by a Dow Jones subsidiary Tradeline.com (formerly IDD
Information Services and GARI Software Associates).

Following IDD Information Services’ name change to Tradeline.com, the
RankDex technology was licensed back to Robin Li for the famous Chinese
search engine Baidu. Since the license’s expiration on 2007-12-09,
RankDex was re-acquired by the IDD Information Services, now owned by
GetUpdated, Inc. Following the appointment of Clinton Cimring, a search
engine optimizer (SEO), as CEO of IDDIS, the search technology is being
redeveloped. The projected launch date has been set for January 1, 2012
for the search engine.

As a technology-based media company, RankDex aims to provide the best
way for people to find information. In addition to serving individual
Internet search users, RankDex provides an effective platform for
businesses to reach potential customers.

“Our mission is to provide the best way for people to find information.
To do this we listen carefully to our users' needs and wants. Have we
collected all the web pages they want to see? Are the pages current and
up to date? Are the search results closely related to their queries? Did
we return those search results instantly? To improve user experience, we
constantly make improvements to our products and services. Our users
definitely notice the many little things that we do differently to
ensure a simple and reliable search experience every time.”
-Robin Li, Founder of Baidu

In addition to serving individual users, RankDex also serves as a media
platform for online marketing customers. They not only provide their
customers easy access to one of the largest online audiences in the US,
but also targeted groups with defined interests as indicated by queries.
Unlike traditional online advertising services, which charge by flat
fee, RankDex’s marketing products and services are performance based.
According to the Dr. Cimring, their Pay for Performance model previously
used under GetUpdated's Search Engine Partner from 2006-2010 has taken
the market “by storm” because it is cost effective and measurable.

“We know that a lot of interesting things are going on in the Internet
space, but we don't want to lose focus. The US’s Internet search
industry is only a newly discovered territory. We see vast untapped
grounds in our home base and we believe there are still plenty of prizes
to be claimed by the best players.”
-Robin Li Founder of Baidu
/Quote

References:

Kowalski 1997 - Information retrieval systems: theory and implementation
 By Gerald Kowalski ISBN 0792399269, 9780792399261

/body