Since college, passion for mathematics has transformed into daily Product Engineering solutions. Certainly, I miss that Math world, but staying happy with parallel world of Engineering. Yes, I think, both help in solving a problem-space by computational; engineering it’s about solving practical problem and math it’s about proving theorems.
Yesterday, stumbled upon YouTube presentation by known scientist Stephen Wolfram who has been involved in huge project of Mathematica and written a book. Recently, he has gone ahead to answer the problem of organizing/computing world’s information.
Wolfram|Alpha is very fascinating project to create first computational knowledge engine to do sophisticated computations, both pure computations involving numbers or formulas, and computations applied automatically to data called up from its data sources. This answers very computational requirement to turn generic information into specific answers where present search engine has struggled.
When we develop software, we also deal with this info but these IT data is very structure as it is directly meant for that targeted system. There is still no system which can read and comprehend any data and make sense out of it.
I was more interested in the internal organization of his answer to semantic search. In this presentation, he describes his system in four major components.
First building block is “Curated Data”, a data set which is derived from different data sources and mapped into structures like XML/RDF. Second is “Algorithm Computing”, a set of methods based on mathematical formulas to drive a scene out of this Curated Data. Third is “Linguistic Analysis” which interprets the human inputted question/queries and map it into one of the appropriate Algorithm to compute. And Final one, “Automatic Presentation” to judge the user interface required for a given query.
Wolfram gave his previous experience with Mathematica to answer open source. In past, he has made huge effort to put the code of Mathematica but he sadly admits there was hardly any interest due to its complexity. I think he is right, this work may not be easily comprehensible at this stage, but may be in future, a better language will represent these scientist code in simpler fashion.
Wolfram way of monetization is very similar to Google propagated Adverts, relevant Ad per query. He also pointed out that he has been successful to bring collaborations with lots of major industry players. He accepts there is still lot of work to be done in machine’s Cognitive Reasoning to normalize data issues like juxtaposition and anthology queries.
As a whole, I still believe, Wolfram can’t be a Google Killer but it can augment “expert answering” space where Google has not affected yet. At the end, Google should either catch up to it, or least, gulp & burp on it.