Case study on privacy and transparency in data publishing

Talk in Portuguese.

Abstract

Technological advances in recent decades have led to the development of techniques for the collection and mass processing of data, including from individuals. This accumulation of information allows governments to fulfill their duties when distributing resources to areas and populations most in need, or even for private companies to develop new products. Therefore, it is necessary that the provision of information be made in a clear and reliable way, or transparently.

However, it is also the responsibility of data holders, whether they are governments or private companies, to ensure that information is protected from misuse, including the re-identification of individuals. Aiming for transparency, past microdata publications were disastrous, despite attempts to keep the identities of individuals whose information found in the databases private. Today it is known, for example, that de-identifying data, that is, removing information such as name or unique identification codes, does not guarantee privacy.

In this way, several anonymization techniques have been proposed and applied in an attempt to balance the right to privacy and maintaining the usefulness of databases. In this lecture, we will address some of these techniques and their advantages and disadvantages, in addition to presenting real cases of individuals who have been reidentified, both in Brazil and worldwide.

Date
17 Feb 2020 13:00 -0300
Location

Instituto de Ciências Exatas,

Universidade Federal de Minas Gerais

Belo Horizonte, MG
Avatar
Gabriel H. Nunes
MSc candidate in Computer Science

Physicist and Computer Scientist

Related