Google pollution

3 January 2006

Om Malik reckons that the geeks are taking over Google. To soak up that quiet time before 2006 really gets going - it looks as though the working 2006 has been postponed to the 4th if out of office replies from the UK are anything to go by - he recommends googling on common first names.

Try Paul, and soon after Paul McCartney you get Paul Graham. The number three entry under Robert is, naturally, Robert Scoble. I tried Chris and Chris Pirillo came out top. Check out the backlinks and soon the reason becomes pretty clear. All of the people cited by Malik are bloggers. Graham's site does not use a conventional blog structure but it's close enough for jazz. It's not so much that geeks are inheriting Google but that famous - that is, heavily linked - bloggers geeks are encroaching on the top pages. Try Jeff and Jeff Jarvis appears at number two.

What do bloggers do? They link to stuff, and mostly other bloggers. I think I packed five such links into the last two paragraphs, so I've done my bit for their already inflated pagerank. Malik's post is only a bit of fun but it does some problems with the Google's results and the disproportionate visibility of blogs.

One is the common belief that Google is some sort of guide to the zeitgeist. Google's creators made a sensible decision to use pagerank to order search results. Citation is, for the most part, a good way of showing how important a piece of information is. It is, also, highly vulnerable to gaming, which is why splogs have been so successful at polluting Google's results. Successive tweaks to the pagerank algorithms deal with the worst abuses as they appear. But there will always be some pollution with a system that depends largely on people "playing by the rules".

The bloggers themselves have unwittingly - and on occasions deliberately - gamed the system by being so profligate with links and made it look as though bloggers are the only community with a voice on the Internet. That is a situation that cries out for a major tweak to the result-ordering algorithms before everyone starts believing that bloggers are genuinely representative of the world population. However, Google has to weigh up whether demoting blogs works for its interests: who else is going to carry Adsense content for the company?