Monday, July 26, 2010
Next step for Wikileaks: Crowdsourcing classified data - Computerworld
WASHINGTON - The release on Sunday by Wikileaks of more than 90,000 documents about military operations in Afghanistan may just be the start of problems for the U.S. government.
The online publication of the documents, which offer an inside -- and potentially embarrassing -- look at the war in Afghanistan between 2004 and the end of 2009, represent a failure by the U.S. to control its classified data from insider threat. And it throws open to the whole world a chance to crowdsource the information the documents contain.
With that in mind, Wikileaks' Editor-In-Chief Julian Assange on Monday urged intrepid researchers to cull the documents for information that the group -- and three publications given access to them -- have yet to uncover. Assange said that Excel, one of the formats in which the material was released, might be the best way to sort through it.
During a news conference that was webcast, he even guided would-be researchers, saying they could use a search term such as "children" to parse the data for casualty reports.
When mining the documents for information, it's important search for something "quite broad...," he said. "Don't tell the data what your prejudices are, but rather let the data tell you what it is."
Now that the secret data has been made public, Assange said he expects academics, students and computer programmers to "come in and do a better job than we have with this presentation."