The buzz surrounding the world of search engines certainly hasn’t died down in the past few weeks. In fact, the people of WolframAlpha have released a package of updates, which according to them is significant, although there really doesn’t seem to be anything beneficial for users at the moment. With this new entry and Microsoft’s recently released Bing, Google may have been feeling left out. Not wanting to share the spotlight, they have announced a series of developments which are not particularly significant, but nonetheless, have served to make journalists write about Google, too.
Google released Google Squared (coincidently, during the launch of Bing ;-)). For the time being, it’s just an application available from Google Labs, but based on the emphasis placed on its presentation, it seems that this is a service in which they are investing important sums of money. What Google Squared is, is probably unknown even to those at Google and I certainly won’t be the one to unravel the mystery! Of course, I did try it out and I shared my impressions and opinions with my colleagues.
The general idea they seem to want to convey is that it is possible to transform unstructured data (meaning information not yet organized in a database) into easily accessible, structured knowledge (basically, the Holy Grail of the Semantic Web). Unfortunately, the outcome is underdeveloped and often, quite useless (it seems that they have accelerated its release for marketing reasons). It is true that in some cases, in which 4 or 5 elements similar to the system are inserted, it is able to provide tables with specific attributes. But, it is also true that when testing things less similar to the examples, the results are unpredictable. It also seems that Google Squared retrieves the reasonable answers mostly from Wikipedia – a data source which is already partially structured, therefore casting doubt on the possibility to apply this approach to content which is truly general.
From a technical point of view, interesting and important developments can be noted, but it is impossible to tell whether they will be useful in the coming months or years. At the moment, it seems to be more of a research project than something that could actually turn out to be a real product.