Privacy Through Randomization
from the interesting dept
IBM is apparently working on a system to protect private data that is being entered into websites. They simply add or subtract random digits to the data that is entered – so that all the data is completely inaccurate. Then, at the end, they correct the aggregate data, so that it’s much closer to reality. The end result is that companies (in theory) get the aggregate data without actually having anyone’s exact personal data. It actually seems like a pretty good idea to me.
Comments on “Privacy Through Randomization”
No Subject Given
This is simply an application of encrypted functions which came out of Berkeley several years ago.