A few months ago, Google launched an experimental new search project, called Google Squared, that literally tries to take all the messy, unstructured information on the Web and put it into neat little, labeled boxes.
It is still very much in Labs, but today it got better. Google Squared can now deal with four times as many squares of data, 120 up from 30. Columns can now be sorted, and results can be exported into Google Spreadsheets were the data can be manipulated, charted, and so forth. In other words, it is turning random facts found on the Web into data that can be played with and computed.
While Google Squared is much better, it is nowhere near ready for mass consumption. If you do a search for “planets,” for instance (see screenshot), it fails to identify Mars in its grid. The first result is Pluto, which officially is no longer a planet. While I too am still resisting the deplanetization of Pluto because of my emotional attachment to it, Google as a cold-hearted, just-the-facts-ma’am search engine doesn’t have the same excuse. And it is not just Pluto, it also lists Ceres (another planet also-ran), Jupiter’s moon Io, and the Asteroid Belt (which most definitely is not a planet). Mars definitely needs to go in there before the Asteroid Belt.
Is this the best Google can do against Wolfram Alpha and other newfangled search engines which also take advantage of structured data to present a deep set of facts for every query? Search for “planets” on Wolfram and it correctly identifies all eight (minus Pluto) including Mars.