I am eagerly waiting for this new search and compute engine promised by Stephen Wolfram. They call it wolfram|alpha (If google always went with the beta release, Wolfram is going even early).This, if it work in the promised lines is going to rock the Internet evolution. From the outset, this is not just a search engine. It is kind of an intelligent searcher who can loosely understand the human requirements.
For long, it was perceived that a search engine driven by natural language processing is the way forward. But it is pretty hard to build such a system since natural language processing is no mean business. Wolfram’s idea is to create an abstraction and then algorithm of these realizable models. Once we can do a mapping of the requirements to algorithm that is computable, at least in principle we can build such a system. But that is a whole lot of heavy statements already. How easy it is to build all these methods and models into an algorithmic framework? He is using the New Kind of Science (NKS) armoury to realize that. We have to wait to get the full rainbow, but when he promises we can confidently expect something big.
Now once the algorithmic mapping (and implementation) is done, then the question of natural interacting between humans and the system comes. Natural language is the way, but according to him we don’t have to worry about doing that as such. Once the knowledge of the individual is made into a computational framework, then that is enough. I am not an expert in this natural language processing and NKS framework, but for sure this is pretty exciting,both from an algorithmic point of view as well as a practical MontBlanc. As Wolfram himself pointed out Pulling all of this together to create a true computational knowledge engine is a very difficult task. Indeed it is still being considered a difficult problem, both in academia and industry. So there is excitement aplenty in the offing. I am eagerly waiting for this to hit soon.
Considering that, the big wig search engine houses including Google are still struggling to make that dream natural language engines (the many pseudo ones in the market are not quite approved). I remember http://www.ask.com started their business in those lines, but never seemed to have crossed that elusive mark of acceptance, atleast not to an extend to capture a world wide wow! If Wolfram has a new way to get this through, that will be a big breakthrough. I cant wait to see that. Wolfram promises that it is going to be very soon. He says it is in May 2009. My guess is that they will release it on May 14,2009.