For several months now, there has been speculation about the possibility that Google will change the PageRank algorithm for a new algorithm that would filter all spam from search engines or at least try to neutralize it. Find out in this article everything that is known so far about the new algorithm, called Trust Rank.
Search engine spam has become a real problem for Google and the other search engines. Unscrupulous webmasters have created thousands of pages with the sole purpose of linking to their main pages and thereby increasing their PageRank (in the case of Google) or any other algorithm used by search engines that take into account the number of links that recommend a page when determining its importance and placing it in a higher position in search results.
This is the technique used to ensure that when you search for “Thieves” on Google, the SGAE website appears in first position.
Google's new algorithm would prevent this type of practice.
The purpose of PageRank is to assign a numerical value to web pages based on the number of times they are recommended by other pages and the PageRank that these pages have in turn. In other words, it establishes the importance of that web page. Its logic is the following: if a web page links to another page, it is because it is recommending it. And if it recommends it, it must be important in the field of the topic that the first web page deals with. A recommendation that comes from a page that is also highly recommended is worth more than a recommendation that comes from a page that almost no one recommends.
Google wants us to find pages of a certain relevance at the top of the search results, which are being recommended by other pages that are also relevant. To determine the PageRank, Google analyses the number of links coming from other websites and their PageRank. The Trust Rank is based on the same basis. But instead of evaluating the importance of a recommendation based on the PageRank of the page that recommends it, it does so based on a series of web pages that have been considered important by humans rather than by algorithms.
Web pages that humans have determined to be important are considered "seed web pages" and their links are assigned a value. And it is this value that will be transmitted throughout the network.
To illustrate with an example: Let's say we have a seed website "A". "A" will transmit a Trust Rank of 100 to all the websites it links to. These websites, in turn, will transmit a Trust Rank of 99 to all the websites they link to. And these websites will transmit a Trust Rank of 98 to the ones they link to.
To mitigate the degradation of Trust Rank as the seed website moves away from it, a correction has been included in the algorithm that takes into account the number of degrees between the seed website and the website receiving the Trust Rank, without completely cancelling out the distance that separates them from the seed.
According to a Stanford University paper on Trust Rank, just as seed websites will transmit Trust Rank, websites that are considered spam by humans who evaluate websites will transmit negative Trust Rank. According to this paper, certain websites will subtract Trust Rank and when a website's Trust Rank level falls below a certain figure, it will automatically become a website that transmits negative Trust Rank instead of positive. (A link to this paper is included at the end of the article.)
The idea of the Trust Rank seems good, but there are certain issues that need to be taken into account:
Who will be the seed websites?
There are discrepancies in this regard… The Stanford document speaks of an algorithm that will determine which websites can be seeded. In certain forums specialising in search engine positioning, it is stated that government and university websites will begin to transmit the seed Trust Rank, and in other forums it is stated that Google has been hiring people from all countries for months to manually evaluate the most representative websites of their country. This last possibility seems the most plausible, since there are indications that Google has indeed hired web evaluators.
Will reverse spam be done?
Off the top of my head and thinking about the not-too-distant future in which Trust Rank will work, it occurs to me that perhaps the same people who play at making the SGAE appear when searching for the word "thieves", may be able to play at sabotaging web pages, mercilessly linking to them from their spam pages and therefore reducing their Trust Rank, so that they do not appear at the top of search engines.
When will Trust Rank be incorporated into Google's algorithm?
No idea… nobody here can agree. One day, Google releases a statement and informs us that it has already implemented it. What is clear is that it will communicate it to the press and to Internet users. It will undoubtedly mean a great qualitative improvement in obtaining search results, so Google will make people aware of it. I doubt that Google's communications department would let an opportunity like this pass by.
Additional information for those who wish to expand their knowledge:
Link to Stanford University document on Trust Rank: http://dbpubs.stanford.edu:8090/pub/2004-17
Search Engine Optimization Course (which will undoubtedly have to be modified the day Trust Rank is implemented, but which already includes the new indexing system with the Google Site Map Generator): Online Course on Search Engine Optimization. The course is free.
