Quantcast

Google Calculator (Google Agent?)

p. “(extlink)Google”:http://www.google.com now allows you to enter math equations directly into the web search field, and it will automatically spit out results, as well as do unit conversions among other things. For instance, if you enter: *speed of light * 2* then google will return

 the speed of light * 2 = 599 584 916 m / s 

p. This function is in addition to Google’s ability to also differentiate between “(extlink)definitions, addresses, phone numbers, and other data”:http://www.google.com/help/features.html . Admittedly, language processing is an intrinsic part of designing a search service/engine, and further development along these lines serves to enhance Google’s core search ability, but could this be the forerunner to the next leap beyond searching?

p. In many ways, this technology seems similar in concept to the “(extlink)Apple Computer’s Data Detectors”:http://developer.apple.com/sdk/index.html#Apple_Data_Detectors in that a computer would be able to examine a piece of data and based on its syntax, structure, and or context possibly determine if it fits a generic class of objects, which could then have specific actions that might be used upon it. e.g. email address could be identified as such, and in addition to a search could present (in addition/instead of the standard web search):

# A directory profile for that address
# A link to email that person
# A list of Usenet posts authored by that email address

p. Where this really seems to get interesting is when incorporated with ideas proposed by Tim Berners Lee (among others) in a “(extlink)Scientific American Article”:http://www.sciam.com/article.cfm?articleID=00048144-10D2-1C70-84A9809EC588EF21 . If you match Google’s ability to parse language and generate automatic meta data with an increasingly “(extlink)semantic web”:http://www.w3.org/2001/sw/ where the resources the search engine is drawing upon have more and more metadata (either explicit or generated based on context) to put them into context and describe their information, the search engine could morph into a powerful method of navigating through and utilizing that information. Combine this with the Google API, and suddenly Google becomes the enabler for a whole ecosystem of information gathering agents.

p. A second, and almost side benefit of this evolution could be a solution to the problem noted by “(extlink)the Register: Blog Noise Achieves Google KO”:http://theregister.co.uk/content/6/33366.html . A Google that can analyze the context and setting of the data on both sides of the search could probably be taught to improve the results.

p. All in all, I think it will be interesting to see where Google is heading with some of their new initiatives.