Wolfram Alpha: The world’s first computational knowledge engine

Posted by Ian on May 16, 2009 in blog | No Comments

Last night witnessed the launch of the world’s first ever knowledge computation engine. Wolfram Alpha is now officially available to the public to solve all your knowledge computation needs. The brainchild of Stephen Wolfram, Wolfram Alpha‘s ambition is to be the one stop shop for formal knowledge on the web. Built on the architecture of Mathematica, the algorithmic computational acrobatics of the engine pulls source material from a broad array of authoritative sources (census reports, the Oxford English Dictionary, and countless other online databases) to present a simple and elegant interface with the world’s knowledge.

The idea behind the engine is that popular search engine’s such as Google or Yahoo are utilised to discover informal knowledge and broad-based web content based on the tags and page ranks of web content. Wolfram Alpha, however, searches formal knowledge structures to provide formal information to users’ questions.

Designed to answer questions, Wolfram Alpha utilises a free-form natural language user interface, and it makes assumptions about search intent – referencing other possible intentions in the knowledge results. A search for ‘truth’ is initially assumed to be a search for a definition, but one can redirect to explore the ‘general concept’, or a search for ‘existence’ initially returns David Cronenberg’s film ‘eXistenZ‘ which can be redirected to the concept’s definition.

My initial concern in using the knowledge engine was that it was an overly clean and simplistic interface that obfuscated (disguised/obscured) the complex strucutres of knowledge creation. Asserting itself as a distilled source of authoritative information, the engine was elevating itself to the level of authority. The balance this criticism, on every query-response page there is a link to ‘view the source information’. Here one can see all the varied resources utilised to create the knowledge document – a bibliography of sorts. It does not however footnote any single figure of piece of information, so there is a fair degree of trust involved in using the engine that the Wolfram Computation Engine does indeed call upon reputible sources in its knowledge computations.

It is a bold step towards enriching users’ experience of the web, and for now, it’s free. Although being built on theories of organic computational evolution, it is a far cry from Terminator‘s Skynet and sentient computing – though it is a step on that road…

EmailBufferFacebookTwitterTumblrLinkedInGoogle+PinterestSina Weibo

Leave a Reply