The Science of Search - the new kid on the block

For those of us born into the information age, "search" has been synonymous with Google. The company hasn't really been around for eternity though, it was established as a private firm only in 1998, after Larry Page and Sergei Brin started it as students in 1996 at Stanford University. Since then, Google has diversified into advertising, e-mail, SMS alerts, online office suites - you name it. Such is its dominance that Google has 73% of market share in search engines, followed by Yahoo at around 16% and then MSN's Live at about 6%.

Last year in July, there was another search engine that claimed to be "better than Google" and "having the most number of indexed pages". It was christened "Cuil" (pronounced Cool), and started by former employees of Google. Unlike many other search engines however, they clearly stated that they would not store any data regarding user activity, or IP addresses of the people who were using their product. Their home page certainly looks cool (forgive the pun) but the problem with them was that they focused more on the quantity of results rather than the relevance. Searches would frequently throw up much less relevant results than what Google would give us. The site had less than 0.02% of market share and it's virtually flat-lined.

But wait, lo and behold, there's a new kid on the block. In the form of a British Physicist's project by the name of "Wolfram Alpha". Strictly speaking, this isn't a search engine per se. What it is is an 'answer engine' or 'computational knowledge engine' as the creator puts it. Where Google returns a list o pages relevant to the search phrase, Alpha strives to solve the problem and present you with an exact answer to your query. Hence while Google will not comment on the reliability of the information that its results refer you to, Alpha makes some judgement calls on the quality of the sources, it supposedly cites academic documents and sources for each result. Stephen Wolfram has announced that this product will be released to the public on May 15th and going officially live on 18th. Mind you, he's already said that its "far from finished".

Wolfram Alpha is built on Wolfram's earlier flagship product, Mathematica, which encompasses computer algebra, symbolic and numerical computation, visualization, and statistics capabilities. The trick with Alpha is knowing how to enter the data so as to get the best possible result. Entering "how much did titanic earn at the box office" would give you a list of results on Google, but what Alpha would do is tell you that the answer is approximately $1.8 bn, according to so-in-so sources.

Obviously, there are also problems associated with this engine, namely that it focusses on raw data more and that its strength lies in Mathematical and scientific calculations and queries rather than entertainment and sports related stuff. WolframAlpha is the brainchild of Stephen Wolfram, 49, a physics prodigy who earned a Caltech Ph.D. at age 20 and won a MacArthur Foundation "genius grant" at 21. He has employed a number of experts to go over the data and hence "curate" it for the search engine to make better decisions. Apparently the product can even tell you the number of people in the US with a particular name. It remains to be seen whether this idea will even dent Google's share, though Google seems to be completely dismissive of the idea.

Google, meanwhile, has announced "Google Squared", a service that - similar to Alpha - will produce charts and graphs of the data you asked for.

The San Francisco Chronicle described the feature in a bit more detail:

[Google Squared] compiles details from several Web pages and organizes them into a table on a single page, with multiple columns like a spread sheet. A search for “small dogs,” for instance, returns a list of breeds, an accompanying image and a brief description, plus the average height and weight of each breed.

It is to be released as a Google Labs Feature sometime around the launch date of Wolfram Alpha. By the looks of it, Google will be steamrolling these guys too, what with Wikia Search ( a 'crowdsourcing' search tool ) already having gone the way of the Dodo. Wikipedia's founder, Jimmy Wales announced the end of the Wikia Search project earlier this year, saying it was no longer viable (he pulled the plug when it came to finances - maybe he needed the money for another expensive dinner with a few bottles of wine at a Florida steakhouse ;-) ).



With the hunger for data, the thirst for knowledge being an omnipresent characteristic of the human race, search engines will always flourish, though one of them will still lead the pack for a long time to come.