Googlebombs No More

A post at the Google Webmaster Blog reports that Google has at-last added a new algorithm to handle googlebomb problem. It's a welcome addition, though Google should have taken care of this years ago.

For those of us who don't remember what's a GoogleBomb : it's a way to manipulate google ranking system (or for that matter any search engine ranking system) by creating lot of links with a wrong anchor text. For example, one of the most popular googlebomb used to show George Bush bio at WhiteHouse website whenever somebody searched for 'miserable failure' (Yahoo or MSN search still shows that if you want to check :)

Many of us are more interested in knowing how Google handled this thing automatically. It's not straightforward as links and anchor texts are an important part of google's search engine ranking system. My guess is that it's somehow based on the sudden appearance of hundreds of new links. Something like this:

First detect the scenarios where there are lot of links to a web page having anchor text not-existing on the web-page itself. If the frequency of such links goes above a certain threshold to affect the ranking of the page for that text, run the googleBomb detection algorithm.

GoogleBomb detection algo will then check two things:

  1. If there is a sudden appearance of lot of links with similar anchor text after certain time t.
  2. If there exists a time t1 (< t) when there was mostly no link with that anchor text

If both is true, then it's a GoogleBomb. Of course, real challenge lies in finding the right threshold values of all the parameters involved -- e.g., what do you mean by a sudden appearance; how many sudden links are required to trigger googlebomb check etc. But Google has lot of interesting data to train their algorithms on -- so it may be a straight-forward analysis for google folks.

I will be happy if Yahoo and MSN search also takes action to defuse such GoogleBombs in their search results. Search is now the most important element of web surfing, and it should be bug-free as much as possible.