This article is more than 1 year old

Google buys search engine – PageRank™ RIP?

Got bots?

Google has bought Kaltix, a three-month-old, three-man Stanford startup that's working on personalized and context-sensitive search. Despite its battalion of PhDs, Google isn't too proud to acquire external search technologies, and earlier this year bought Applied Semantics for its CIRCA ontology, which "understands, organizes, and extracts knowledge from websites and information repositories in a way that mimics human thought".

Google has made no secret of its goal to "understand" the web, an acknowledgement that its current brute-force text index produces search results with little or no context. The popularity of Teoma demonstrates that even a small index can produce superior results for certain kind of searches. Teoma leans on existing classification systems.

While Google relied on PageRank™ to provide context, all was well. But PageRank is now widely acknowledged to be broken, so new, smarter tricks are required.

Regarded as heresy when we raised the issue last spring, now some of Google's warmest admirers, MetaFilter's Matt Haughey and web designer Jason Kottke have acknowledged the problem.

As Gary Stock noted here last May, Google "didn't foresee a tightly-bound body of wirers. They presumed that technicians at USC would link to the best papers from MIT, to the best local sites from a land trust or a river study - rather than a clique, a small group of people writing about each other constantly. They obviously bump the rankings system in a way for which it wasn't prepared."

Although it's tempting to suggest that bloggers broke PageRank™ it might equally be the case that the Blog Noise issue is emblematic rather than causal. Blog Noise - in the form of 'trackbacks', content-free pages and other chaff - is the most visible manifestation, but mindless list-generators are also to blame for Google's poor performance. And the truth is every successful search engine will find itself engaged in an arms race with gamers. (Deliciously, in the case of email spammer Elwyn Jenkins, a former e-currency salesman who proselytizes weblogs by day, and by night offers advice on how to improve your PageRank, the bloggers and the Google gamers are one and the same [includes screenshots]).

Daniel Brandt, who runs the 100,000-document NameBase archive, has been PageRank&trade's most severe critic, and acknowledges that it lives on in name only. Google no longer performs a monthly recalculation of PageRank values and anchor text is the most highly valued criteria for a search, he says. Which makes Google hardly distinguishable from AltaVista five years ago.

"Quantity not quality is the word on the street," he told us. "The old method of doing PageRank™ which had more integrity or consistency, to it is no longer being done. PageRank™ was a bad idea philosophically to begin with, but now some spammer can set up hundreds or thousands of sites automatically with anchor texts pointing to one page. Before, each would have such a tiny PageRank™ that it wouldn't amount to a hill of beans. Now you can do that and if the anchor text is carefully chosen it will make a difference," he reckons. "The cure is worse than the disease".

As an example he cites the results for the search "discount brokers" - of all the discount brokers, Google's top results is a page which has been empty for a year.

So perhaps Google needs to give a formal burial to PageRank™ rather than maintaing
ing a goulish afterlife as a marketing gimmick. The future promises to be much more interesting. ®

Related Links

"Make Money Online" - How bloggers game Google
PageRank is Dead [Zawodny]
"Google is busted" [Kottke]
Maybe I should write for the Register UK too? [Haughey] - follow-up

Related Stories

Google to fix blog noise problem
Blog noise is 'life or death' for Google
Google - the only archive we'll ever need?
Google heals the sick

More about

TIP US OFF

Send us news


Other stories you might like