Beyond PageRank

In their WWW2006 paper, Beyond PageRank: Machine Learning for Static Ranking, the authors Matthew Richardson, Amit Prakash and Eric Brill, all from Microsoft, demonstrate that for static (query-independent) ordering of Web pages, machine learned link independent page features perform significantly better than Google's PageRank (tm). But enough with the geek talk, I just wanted to highlight the following paragraph from the introduction:

Google is often regarded as the first commercially successful search engine. Their ranking was originally based on the PageRank algorithm [5][27]. Due to this (and possibly due to Google's promotion of PageRank to the public), PageRank is widely regarded as the best method for the static ranking of Web pages.

Though PageRank has historically been thought to perform quite well, there has yet been little academic evidence to support this claim. Even worse, there has recently been work showing that PageRank may not perform any better than other simple measures on certain tasks.

Perhaps more of a rimshot than a potshot... You can read the rest of it here. Good stuff for those so inclined.

Spread the dishum:  digg it del.icio.us reddit furl My Web

- June 14, 2006 5:59 PM // Technology