Date Mining in a Haystack

Data mining is the process of collecting vast amounts of information about people, generally from the internet, and using that information for some purpose. The most commonly known uses are credit card companies using the process to identify stolen cards. Many online companies data mine as well, using the gathered information to target ads to people depending on their tastes. Both of these uses seem rather handy, although the ad targeting can get annoying at times. However, not all data mining is so friendly. the department of defense of the U.S. government uses data mining to identify possible terrorist threats. In a article on Wired, Bruce Schneier points out how ineffective the process is and how much of our privacy is lost for so little gain. Schneier’s point is that because of the complexity of attributes that lead to terrorist attacks, no data mining system would be able to accurately predict any attacks in the foreseeable future – reasonable if one thinks of the internet as a haystack and the key information as a needle, broken into pieces and thrown in. This means the citizens of the U.S. are essentially being watched by their own country for no security gain. This was proven by the NSA’s eavesdropping program: it gave thousands of tips per month, which were followed up by the FBI and each one was a dead end, costing thousands of man hours which could have been spent actually making the country safer. Schneier is not totally against date mining: he supports the uses by credit companies to fight fraud, by sites like Amazon and Google for showing him things he might buy, all because there is a obvious gain for the loss of privacy.

Personally, I agree with Schneier on data mining. If the uses it is being put to work, like the credit card and online shopping, and I benefit from it then I support it. But the current projects attempting to protect us do sound to much like a Big Brother government in the making. As the system currently just tells the government everything I do on the internet but fails to catch any terrorists, and very few other criminals, I do feel it is logical to keep using it. If it worked, if this system had stopped one terrorist attack, I would feel better about its continued activities, but seeing how it has not managed that, I feel that it is not worth the lose in privacy. Security Data Mining on such a large and random scale as terrorism needs serious work, and until it can actually effectively work and protect us, I don’t feel like sacrificing my privacy, on principle. It really comes down to Ben Franklin’s paraphrased quote: He who gives up freedom for safety deserves neither. And giving up privacy is to close the freedom for me.

Bruce Schneier is a cryptographer, computer security specialist and writer. He maintains generally liberal views on cyber security and disclosure.