Open Access News

News from the open access movement

Saturday, May 16, 2009

Launch of Wolfram|Alpha

Wolfram|Alpha has officially launched.  But don't expect to try this exciting OA "knowledge engine" any time soon.  User traffic has brought it to its knees. 

You can still read the project FAQ and blog or watch the overview video (recommended).

Alpha is open in several senses.  It's free to use.  It has an open API.  And it draws upon a combination of open and closed data, though in most cases (even for open data) using Wolfram-hosted, curated, and refined copies of the original datasets.  BTW, Alpha not only draws upon many data sources, but cites its sources in its answers.

Alpha is built on nearly 6 million lines of (Wolfram) Mathematica code.  That code is not open source, but Alpha provides the Mathematica source for its outputs.  From the launch announcement:

...[Y]ou can take the output from Wolfram|Alpha and actually compute with it in Mathematica....[Y]ou can also click the Use Input button on particular Wolfram|Alpha output pods in the notebook, and get Mathematica input that would generate the basic output in the pod....

One way to see the Mathematica code is to get the Live Mathematica Notebook.  But if you just want to know the Mathematica code for a particular Wolfram|Alpha output pod, you can click on the pod and look at the popup.

So long as there’s a simple representation of the pod computation in Mathematica, there’ll be a “Mathematica plaintext input” field in the popup —that you can copy and paste anywhere....

See our past post on Alpha and some of the early coverage and reviews.

Comment.   Alpha is orders of magnitude better than any previous attempt to go from user queries beyond lists of web pages to direct, composed answers.  It reminds you why computers are cool.  It will be terribly useful in every area it covers, and will steadily enlarge the number of areas it covers.  It's very exciting even from a narrow OA perspective because sophisticated Q&A services are among the most effective ways to solve the last-mile problem for knowledge.