Physicist Stephen Wolfram says the goal of his free Wolfram Alpha program, which will be available to the public starting in the middle of May, is to “make expert knowledge accessible to anyone, anywhere, anytime.” Wolfram’s computational knowledge engine was recently demonstrated at Harvard University’s Berkman Center for Internet and Society. It is designed to answer questions directly instead of retrieving Web pages in response to queries. Wolfram Alpha employs natural language processing to enable the use of normal, spoken language queries by users, and Wolfram says the program has addressed many of the challenges of interpreting people’s questions. The program computes many answers on the fly by capturing raw data from public and licensed databases, along with live feeds. Wolfram says that trillions of pieces of data were selected and managed by a team of experts at Wolfram Research, and that these experts also tweak the information to ensure that it can be read and displayed by the system. He says the system has become proficient at eliminating “linguistic fluff,” or words that are unnecessary for the location and computation of relevant data. This statement disappointed Boris Katz of the Massachusetts Institute of Technology, who is head of the Start natural language processing project. “I believe [Wolfram] is misguided in treating language as a nuisance instead of trying to understand the way it organizes concepts into structures that require understanding and harnessing,” Katz says.

View Full Article

For more information please visit: http://www.cpccci.com