|
IBM Researcher Brings Analysis to Enrcypted Data |
|
|
|
IBM announced today that a researcher had cracked a long-standing cryptographic challenge, using what is termed "fully homomorphic encryption" to allow data mining and analysis of encrypted data without revealing the sensitive data itself.
While the release touts the breakthrough as a possible aid to cloud computing, questions arise as well: If data is sensitive to begin with, isn't analysis of that data equally - or more - sensitive? While hackers want names, SSNs, and credit card numbers that would, in this scenario be less exposed, many others would welcome insight to data reports about a company's customer base - geographic densities, gender and age breakdowns, churn rates, revenue per customer, and much more.
Would you entrust those reports to a third party any more than you might customer data? The breakthrough appears very promising in theory, but how can it be used on a practical basis?
Only registered users can write comments.
Please login or register. |
