2017-2018 exchange Sept Oct 2017 Newsletter FINAL | Page 13

From a records management perspective, it appears that big
data can cause big problems. Simply put, for big data propo-
nents, more is more while records managers are looking to
have records and non-records systematically deleted and avoid
over retention issues.
A potentially significant issue is if data is extracted in such a
manner that a record can be recreated. This is particularly prob-
lematic if the record at issue is a policy/account record contain-
ing personally identifiable information.
For privacy reasons, pseudomization is used as
a technique to minimize the risk of being able to
identify an individual through the data that is
used. The technique consists of replacing identi-
fying fields with pseudonyms (e.g., John Smith=
Q23O). The flaw with this approach is the exist-
ence of the look-up table, which enables re-
identification.
Anonymization techniques can also be used to protect the iden-
tify of individuals. Unlike pseudomization, identifying fields (e.g.,
date of birth, names) are irreversibly destroyed.
13