KDnuggets : Polls : TIA effectiveness and privacy
Poll
DARPA says data mining in TIA initiative can be effective (find likely terrorists with few false positives), and preserve privacy of innocent citizens. Do you think TIA will be [81 votes total]

Effective, and preserve privacy (10) 12%
Effective, erode privacy, but is worth it (16) 20%
Effective, erode privacy, but not worth it (12) 15%
Ineffective (32) 40%
Don't know about TIA (11) 14%

Comments

Peter Fitton, Ineffective
I voted "ineffective." I expect it to be a lot of money spent with little to show for it.
This is not like screening credit card transactions for fraud, or predicting which cellular customers will churn. There are too few needles (terrorists) in too large a haystack (the population of residents of and visitors to the US) to get anything valuable using current hardware and data mining techniques. Furthermore, the data integration effort that forms the foundation of the project will turn out to be much more difficult than now envisaged.

Chris Clifton, CAN be ...
The key word here is CAN be - I am confident it is possible develop technology such that TIA will not significantly erode privacy, and will be at least somewhat effective (a substantial improvement over today's systems.) WILL be is another issue - I don't know enough about what is being done to have confidence that the program will meet its (very ambitious) goals.

Alessandro ZANASI, Effective, and no risk.
I answered: Effective, and preserve privacy. The TIA enabling technologies are already taught at universities, in use in the business sector (e.g.: email analysis, feelings extraction and competitive intelligence through text mining...) and in national security field (with some applications more advanced than the commercial ones). These experiences were considered so effective that TIA has been proposed. I personally don't see more risks for the common citizen privacy than before TIA implementation.

Will Dwinnell, Effectiveness and Privacy
I voted for "Effective and erode privacy, but worth it", but... Alot will depend on the details of course, but I believe that a large technical information effort could be at least moderately efective at preventing or deterring terrorism.
Depending on the "privacy" being given up, it might or might not be worth it.

Josh Froelich, Remember the Carnivore
I just wanted to reference the governments earlier use of the carnivore program, where isps were asked to install email monitoring like software so the fed could network search parts of email content in logs for keywords like "terrorism". I remember watching how a congressional committee reacted to the privacy implications. It was not positive. There is definitely the issue of technical literacy, as was apparent from how hard it was for the committee to understand the fine line between aggregate profiling and intruding on a single individual's private life. This communications barrier shows that even if it could be effective, it seems most non-technical people would be unable to understand its implications and assume the worst.

Editor, TIA and Data Mining: Effective, Preserve Privacy? DARPA recently released to Congress its report on Terrorism (formerly "Total") Information Awareness program (TIA). DARPA says TIA is now a research program but when deployed, it can be effective - find likely terrorists while keeping false positives to a manageable number. Critics say it will be ineffective and will erode privacy. What do you think?


KDnuggets : Polls : TIA effectiveness and privacy

Copyright © 2003 KDnuggets. Subscribe to KDnuggets News!