|
ABSTRACT:
|
The problem of removing sensitive information from data before it is
released publicly, or turned over to less trusted analysts, underlies
much of the unwillingness to share data. The solution is to sanitize, or
deidentify, parts of the data. When dealing with network addresses, the
set of available addresses is finite. This limits some aspects of the
sanitization. We analyze this problem in detail, and suggest approaches
to ameliorate it
|