Google Pagerank Algorithm Explained

Google uses the PageRank algorithm to rank pages in their search engine results. The algorithm is based on analysing link structures: each link can be seen as a vote by a page for another page it links to. It is necessary to have a basic understanding of how PageRank works to be able to optimize web sites for the Google search engine .

History of PageRank

PageRank was invented by Larry Page at Stanford University, and named after him. Page began the project in 1995, and worked on it with Sergey Brin. The project then lead to the prototype Google in 1998. Google Inc. was founded to manage the Google search engine, which used PageRank. Currently the trademark PageRank belongs to Google Inc., but the original patent for the PageRank algorithm is assigned to Stanford University.

Google has since grown to the largest search engine in the world, with nearly two thirds of searches made on Google. The PageRank algorithm is still one of the factors taken into account in Google's search engine result ranking, and it is constantly followed by interested search engine marketing specialists around the world.

The PageRank Algorithm

A good analogy for the PageRank algorithm is that each link is a vote for the target page. This link structure of the web is analyzed recursively, so that the value of each vote depends on the PageRank of the voting page, calculated from the votes in the previous iteration of PageRank.

When a new page comes into being in this link structure, it will only get a PageRank value once other pages link to it. Each link will give some PageRank to the new page, the amount a single link gives depending mostly on two things: the PageRank of the linking page, and the number of links on it. The more outbound links a page has, the less weight each of them will have.

The actual algorithm used by Google is, naturally, more complicated then this. That algorithm also takes into account page topics. Pages with the same, or related topic that link to each other will carry more weight with their links than completely different pages that link to each other.

It is thought that raw PageRank data exists as a floating point number for each page. These real time values are constantly updated by Google crawlers. The visible PageRank value that can be seen on the Google Toolbar is on a scale from zero to ten. This is exported from the real time PR data approximately once every three months. On Google Directory the PageRank is reported as an eight unit measure, which is also periodically compiled from the real time data, though less often than the toolbar PageRank is.

In Conclusion

The voting analogy describes the PageRank algorithm fairly well, but, though Google may claim so, it is not a completely democratic system. The weight of each link depends both on the PageRank of the page linking and the number of links it has. To improve PageRank, more inbound links, preferably from pages with high PageRank, must be acquired.