Wolfram Alpha and Computational Search

I was honestly more excited about the launch of Wolfram Alpha (http://www.wolframalpha.com/) than Microsoft Bing in May this year. It must be noted that Wolfram Alpha is a computational engine and the service that Wolfram provides is very different from Google or Microsoft search.

Stephen Wolfram, the founder of Mathematica led the effort to devise a strategy that enables the Wolfram Alpha engine to understand the language and interpretive intricacies of the web as compared to the more typical semantic tagging approach. Though not on the same scale, Wolfram's efforts in the present day are very similar to an ancient Sanskrit grammarian named Pāṇini (पाणिन) from the Indian subcontinent who lived around 27 B.C.E and produced 3,959 sutras or rules of  the Sanskrit grammar. Pāṇini essentially came up with formal rules to describe Sanskrit grammar and used these rules to compute poetic forms in Sanskrit. Wolfram's task of attempting to interpret the "language of the web" and making all available data computable by leveraging models, is much more arduous in terms of scale and this is why I am fascinated by the efforts of Wolfram Research.

The best way to understand how the computational knowledge engine differs from existing search engines is to run a few queries . As an extremely trivial example, I typed in "Australia India" and the below results were returned in Wolfram Alpha. A similar query in either Google or Bing led to a whole bunch of search engine results that were not relevant (purely) in context of quantifying comparative (computational) statistics of the two countries.