Web impact excluding link farms and other black SEO practices

[Isidro F. Aguillo]

The second 2013 edition of the *Ranking Web (Webometrics) of 
Universities*, including more than 21000 higher education Institutions 
has been published:

*http://www.webometrics.info/*

The rank is based on a composite indicator that combines webometric and 
bibliometric variables, the first ones collected from the Web during the 
last July while the second one is the updated Excellence indicator (the 
10% most cited papers in 21 disciplines) for the years 2007-2011 (SIR).

Major innovation is a new method for calculating web visibility or 
impact, the most important variable (50%) for the composite indicator 
used for building the Ranking. The former method consists of combining 
the number of domains originating the inlinks to the university 
webdomain (backdomains) and the square root of the total number of those 
inlinks: sqr(backlinks). In this way it was possible to grant proper 
recognition not only to most popular web contents or most prestigious 
institutions but also to the diversity and strength of the impact sources.

Unfortunately a few webmasters choose the easy way to obtain additional 
link visibility contracting external services of link farms or creating 
their own ones by forcing scholars and students to overlink to the 
university webdomain from external blogs or similar social tools.

In the current edition the top 10, an arbitrary number that is open to 
future modifications, sources (domains) of links and all the involved 
links are excluded from the impact indicator. For most of the 
universities this action exclude student blogs, sports teams pages, 
non-academic forums, alternative domains and mostly local sources of 
links, so true global impact is left as main component. Obviously bad 
practices are strongly penalized with the new system. In a few cases 
hacked websites due to poor management of the webmasters are being 
overlinked for more than ten domains, including nasty sources like 
pornographic, fake-products selling or hate pages. We are identifying 
these individual cases for exclusion.

On the positive side, we expect this change will promote 
internationalization of the contents. Webometrics is not only ranking 
websites but the overall performance of the universities. In that sense, 
it has a clear advantage on other similar rankings where results are 
obtained only after long-term efforts. A strong and well directed web 
policy involving not only ICT departments but everybody else in the 
university is going to have a huge impact in the Ranking.

Current leaders by region are:

*North America*: Harvard, MIT, Stanford, Berkeley
*Latin America*: Sao Paulo, UNAM, Campinas, Buenos Aires
*Europe:* Oxford, Cambridge, UCL, ETH Zurich
*Eastern & Central Europe:* Lomonosov Moscow State University
*Asia:* National Singapore, Tsinghua, Tokyo, National Taiwan
*Africa: *Kwazulu Natal, Cape Town, Stellenbosch
*Oceania:* Australian National University, Monash, Melbourne, UNSW
*Middle East:* Tel Aviv
*Arab World:* King Saud
*South Asia:* IIT Bombay


-- 
******************************

Isidro F. Aguillo, HonDr.
The Cybermetrics Lab, IPP-CSIC
Grupo Scimago
Madrid. SPAIN

isidro.aguillo at csic.es
ORCID: 0000-0001-8927-4873
ResearcherID: A-7280-2008
Scholar Citations: SaCSbeoAAAAJ
Twitter: @isidroaguillo
Rankings Web: webometrics.info

******************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20130904/32ca8e44/attachment.html>