Google offers more of the same

By M H Ahssan

The hype this week in tech circles has been Google's caffeine addiction or, for the more technically minded, its next-generation search technology. The "caffeine" code-name refers to a new version of the world-dominating search engine that the company claims will be faster and more accurate.

Most users will not notice the difference, as the results page, format and layout will be unchanged. The only major difference with the new system is that it will scour more of the web for the search terms entered.

Statements on the company blog have elaborated a little but have left most of the juicy bits out: "For the past several months, a large team of Googlers has been working on a secret project: a next-generation architecture for Google's web search. It's the first step in a process that will let us push the envelope on size, indexing speed, accuracy, comprehensiveness and other dimensions."

Despite Google's claim, this probably won't help the majority of websites, which will remain buried on page 57 while the likes of Wikipedia and huge conglomerate travel or merchandise affiliates remain in that elusive top 10 on the first page - and that's without the ones that pay Google to be there. Google's Page Rank system, named after co-founder Larry Page, naturally favors larger websites with more inbound links - they are seen by Google as more relevant and important. Actual content is less of a ranking factor.

Over the years wily webmasters and spammers have developed countless tricks to deceive Google into listing their websites as more relevant and gain higher placements; this in turn deceives readers into clicking on the link and visiting the site.

The latest upgrade may be an effort to prevent this link spamming and underhanded search engine optimization. However, favoring huge corporate sites is hardly delivering more relevant results - it's a tough balance.

In reality, it doesn't really matter if Google indexes 3.1 million pages for the term "Spengler" on the new system (which can be tested at http://www2.sandbox.google.com) compared with 2.1 million on the old, if most people never go further than page one of the results anyway.

The catalyst for this redoubling of effort by Google, which already has more than 60% of the global search market, is Microsoft's latest baby, Bing, and the software company's recent partnership with Yahoo, which has lost out in the search race since the birth of the Google Goliath 10 years ago. Bing has been good news all round as it has provided an alternative, so those looking for real search relevance should try more than just one search engine.

Blindsearch (http://blindsearch.fejus.com) is a good way to test search relevance when comparing the three major engines - Google, Bing and Yahoo. It provides results from the three in randomly ordered columns when search terms are entered. The user then selects which one they feel has provided the most relevant results. It wasn't surprising that Wikipedia and those huge conglomerates featured in nearly all of the searches we tested on it, so the pattern is there with all three major players. Countless websites and resources are still hidden from searchers, and the results we see on the front pages of search engines these days are still the same tired handful of websites that represent less than 1% of the web.

Internet
Chinese government officials stated this week that they will not force the mass installation of Internet filtering software but will be recommending and using "Green Dam-Youth Escort" in schools, Internet cafes and public places. The decision followed global condemnation and fierce resistance from foreign computer manufacturers and the burgeoning population of Internet users in China.

A number of Asian computer companies, including Taiwan-based Asustek and Acer and China-based Lenovo, have been offering the software to consumers but not forcing them to install it. The software, designed by a Chinese company, has been designed to block pornography as a parental control. However, it has always been known that the government have been using it to stifle any politically sensitive websites.

Science
The boffins at the European Organization for Nuclear Research, better known as CERN, have announced the restart of the biggest machine ever made. The Large Hadron Collider, a US$10 billion atom smasher that was destined to probe the origins of the universe last year (see Building a backyard black hole, Asia Times Online August 30, 2008), is set to be fired up again for tests in November following a breakdown that delayed experiments for over a year.

The malfunction was caused by the failure of an electrical connection resulting in a short circuit that caused a helium leak and damaged some of the massive magnets in the tunnel. Engineers have been testing the 10,000 superconducting connections and repairing those with abnormally high resistance.

The 27-kilometer proton race track will run an initial two beams at low power so the scientists can experience running the machine safely while minimizing the risk of a failure similar to that which rendered the beast inoperable last year. Once they are confident, the power will be cranked up a little, but no particle colliding is scheduled until the latter half of 2010. So we can put off "black hole swallowing the Earth" conspiracies until then.