Name: | Description: | Size: | Format: | |
---|---|---|---|---|
645.18 KB | Adobe PDF |
Advisor(s)
Abstract(s)
No mundo digital, toda a atividade humana deixa um rasto de
dados que constitui um recurso cada vez mais valioso, para avaliação e
definição de estratégias nos mais variados domínios. A partilha desses dados,
sendo socialmente importante, implica o respeito pela privacidade individual
e portanto a sua anonimização. As atuais leis e regulamentos sobre
privacidade oferecem orientações limitadas para lidar com um vasto leque de
tipos de dados, ou com técnicas de reidentificação. Este trabalho pretende
ilustrar um processo de anonimização, comparando para vários modelos de
privacidade a perda de informação e a utilidade do conjunto de dados
resultante. Encontrar o equilíbrio entre privacidade e utilidade é um desafio
que pode ser mais facilmente alcançado por quem melhor conhece o
significado dos dados e dos objetivos que se pretendem alcançar com eles.
In the digital world, all human activity leaves a trace of data that is growingly valued for the evaluation and definition of strategies in varied domains. The sharing of those data, being socially relevant, implies the respect for individual privacy and so, its anonymization. The current laws and regulations about privacy offer limited guidance to deal with the vast range of datatypes or with techniques of re-identification. This work aims at illustrating a process of anonymization, comparing to several models of privacy, the loss of information and the usefulness of that dataset resulting from the anonymization. Finding a balance between privacy and utility is a challenge that can be more easily found by those who know better the meaning of the data and objectives aimed at.
In the digital world, all human activity leaves a trace of data that is growingly valued for the evaluation and definition of strategies in varied domains. The sharing of those data, being socially relevant, implies the respect for individual privacy and so, its anonymization. The current laws and regulations about privacy offer limited guidance to deal with the vast range of datatypes or with techniques of re-identification. This work aims at illustrating a process of anonymization, comparing to several models of privacy, the loss of information and the usefulness of that dataset resulting from the anonymization. Finding a balance between privacy and utility is a challenge that can be more easily found by those who know better the meaning of the data and objectives aimed at.
Description
Keywords
Data anonymization k-anonymity ℓ-diversity t-closeness ENADE
Citation
Paula Prata Maria Eugénia Ferrão, Wilson Santos, Gonçalo Sousa. “Privacy Preserving Versus Utility Preserving in Data Anonymization: a study in higher education”, RISTI-Revista Ibérica de Sistemas e Tecnologias de Informação, RISTI, N.º 40, 12/2020, pp.112-127. http://www.risti.xyz/issues/risti40.pdf (in Portuguese)
Publisher
Bertil P. Marques, ISEP, Instituto Politécnico do Porto, PT