Understanding TrustRank Algorithm
You may have heard about TrustRank and yet still not know what it really is. So here’s some information on just what TrustRank is, and what the algorithm is that runs it.
TrustRank is basically a ‘enhanced’ set of link analysis techniques that was first penned/described in a paper by Yahoo! research staff and Stanford University researchers. According to the paper, “TrustRank is used for semi-automatic separation of useful web sites from those that are spam”.
Most ’spam pages’ are created solely with the intention to mislead different search engines and gain rank disingenuously. These spam pages are created mainly for commercial (mainly affiliate) reasons and use different types of SEO and SEM techniques to achieve high rankings on the Top 3 Search Engines or SERPS. Although it is fairly easy for experts to identify spam technically, it is a extremely time consuming and expensive process to manually evaluate each and every page.
One of the most popular methods of improving page rankings on any of the search engines is by artificially increasing the perceived importance of any web document via complex linking tools. Even Google’s PageRank uses similar ways to determine the actual relative importance of any and all web documents that might have been subjected to different types of manipulation.
The TrustRank Algorithm
The TrustRank algorithm is a specific formula used for rating the overall quality of different websites. The TrustRank algorithm was initially published by Gyongyi, Garcia-Molina and Pedersen in 2004. The important idea behind the TrustRank algorithm is quite similar to that of the PageRank algorithm which is, to take the linking structure into account for generating a measure for the type of quality a particular webpage has. This algorithm is considered by many experts as a further development on the existing PageRank algorithm.
The starting point of the TrustRank algorithm lies in the selection of trusted web pages manually. The entire process of trusting pages works on the same lines as in PageRank. The inverse PageRank or negative measure propagates backwards and is the measure for spam or bad pages. For the TrustRank algorithm either of the measures can be taken into account. Here’s the mathematical translation of the TrustRank algorithm:
TrustRank = M-1 * X
Where the matrix M is given by:
M = 1 – d T with
Tij = 1 / Cj (if page j is linking to page i)
Tij = 0 (otherwise)
d is a damping factor and X the source vector of the trust.
The inverse PageRank is given by
Minv-1 * Xinv with
Minv = 1 – dinv Tinv
The inverse transition matrix Tinv is definied by
Tij = 1 / nj (if page i is linking to page j)
Tij = 0 (otherwise)
dinv is again another damping factor while Xinv is considered to be the source vector for the bad or spam pages and nj represents the number of incoming links on page j.
Of course, if you are creating a website then you will not have to worry too much about the TrustRank algorithm. In fact, you can visit http://www.seointelligence.com/ for a Free trial account for implementing strategies which will help in keeping your webpage higher in PageRank and TrustRank.
By Jeremy Hermanns | June 25, 2008 | Filed Under Google, SEO Advice, SEO Tools
comments
Leave a Reply
-
Sneak Preview #1
-
Sneak Preview #2
-
Support
-
