George Duncan is a statistics professor at Carnegie
Mellon University in Pittsburgh, Pennsylvania. He writes in the
journal Science that traditional methods of anonymising people's
database records are not good enough.
He said that databases "de-identify" people by masking important
information such as their Social Security number or their birthday,
but that this does not render them unidentifiable. Anyone who can
access more than one characteristic of a person in a database has a
chance at identifying the person, he said.
The problem is that the very information that most closely
identifies a person is likely to be that in which the organisation
behind the database is interested, he wrote, meaning that it cannot
be deleted or masked.
"The question is, how can data be made useful for research
purposes without compromising the confidentiality of those who
provided the data?" Duncan said in a statement.
Duncan said that it would be possible to build systems which
make this kind of identity reconstruction impossible. He also said
that further user-specific restrictions on the use of information
in databases would go some way to solving the problem.
It is, said Duncan, a difficult problem to solve. "Achieving
'adequate' privacy will require engineering innovation, managerial
commitment, information cooperation of data subjects and social
controls (legislation, regulation, codes of conduct by professional
associations and response to reactions of the public)," Duncan
wrote in Science.
As public and private bodies gather increasing amounts of
information on customers and citizens, concern is growing about
individuals' rights to privacy within such systems. Some opposition
to the Government's identity register is based on fears about the
abuse of or errors in the database that would lie behind any ID
card.
Privacy watchdog the Information Commissioner has warned that
the UK is becoming a surveillance society without adequate privacy
safeguards, and the House of Lords earlier this year said they
would probe the constitutionality of such widespread
surveillance.
