Interactive Anonymization for Privacy aware Machine Learning

Bernd Malle, Peter Kieseberg, Andreas Holzinger

Research output: Contribution to conferencePaper

Abstract

Privacy aware Machine Learning is the discipline of applying
Machine Learning techniques in such a way as to protect and retain personal identities during the process. This is most easily achieved by first
anonymizing a dataset before releasing it for the purpose of data mining or knowledge extraction. Starting in June 2018, this will also remain
the sole legally permitted way within the EU to release data without
granting people involved the right to be forgotten, i.e. the right to have
their data deleted on request. To governments, organizations and corporations, this represents a serious impediment to research operations,
since any anonymization results in a certain degree of reduced data utility. In this paper we propose applying human background knowledge via
interactive Machine Learning to the process of anonymization; this is
done by eliciting human preferences for preserving some attribute values
over others in the light of specific tasks. Our experiments show that human knowledge can yield measurably better classification results than a
rigid automatic approach. However, the impact of interactive learning in
the field of anonymization will largely depend on the experimental setup,
such as an appropriate choice of application domain as well as suitable
test subjects.

Conference

ConferenceEuropean Conference on Machine Learning and Knowledge Discovery ECML-PKDD
CountryMacedonia, The Former Yugoslav Republic of
CitySkopje
Period18/09/1722/09/17
Internet address

Fingerprint

Learning systems
Operations research
Data mining
Industry
Experiments

Keywords

  • privacy aware machine learning

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Malle, B., Kieseberg, P., & Holzinger, A. (2017). Interactive Anonymization for Privacy aware Machine Learning. 15. Paper presented at European Conference on Machine Learning and Knowledge Discovery ECML-PKDD, Skopje, Macedonia, The Former Yugoslav Republic of.

Interactive Anonymization for Privacy aware Machine Learning. / Malle, Bernd; Kieseberg, Peter; Holzinger, Andreas.

2017. 15 Paper presented at European Conference on Machine Learning and Knowledge Discovery ECML-PKDD, Skopje, Macedonia, The Former Yugoslav Republic of.

Research output: Contribution to conferencePaper

Malle, B, Kieseberg, P & Holzinger, A 2017, 'Interactive Anonymization for Privacy aware Machine Learning' Paper presented at European Conference on Machine Learning and Knowledge Discovery ECML-PKDD, Skopje, Macedonia, The Former Yugoslav Republic of, 18/09/17 - 22/09/17, pp. 15.
Malle B, Kieseberg P, Holzinger A. Interactive Anonymization for Privacy aware Machine Learning. 2017. Paper presented at European Conference on Machine Learning and Knowledge Discovery ECML-PKDD, Skopje, Macedonia, The Former Yugoslav Republic of.
Malle, Bernd ; Kieseberg, Peter ; Holzinger, Andreas. / Interactive Anonymization for Privacy aware Machine Learning. Paper presented at European Conference on Machine Learning and Knowledge Discovery ECML-PKDD, Skopje, Macedonia, The Former Yugoslav Republic of.26 p.
@conference{795397df46454061b533ff3eaec2b2b9,
title = "Interactive Anonymization for Privacy aware Machine Learning",
abstract = "Privacy aware Machine Learning is the discipline of applyingMachine Learning techniques in such a way as to protect and retain personal identities during the process. This is most easily achieved by firstanonymizing a dataset before releasing it for the purpose of data mining or knowledge extraction. Starting in June 2018, this will also remainthe sole legally permitted way within the EU to release data withoutgranting people involved the right to be forgotten, i.e. the right to havetheir data deleted on request. To governments, organizations and corporations, this represents a serious impediment to research operations,since any anonymization results in a certain degree of reduced data utility. In this paper we propose applying human background knowledge viainteractive Machine Learning to the process of anonymization; this isdone by eliciting human preferences for preserving some attribute valuesover others in the light of specific tasks. Our experiments show that human knowledge can yield measurably better classification results than arigid automatic approach. However, the impact of interactive learning inthe field of anonymization will largely depend on the experimental setup,such as an appropriate choice of application domain as well as suitabletest subjects.",
keywords = "privacy aware machine learning",
author = "Bernd Malle and Peter Kieseberg and Andreas Holzinger",
year = "2017",
language = "English",
pages = "15",
note = "European Conference on Machine Learning and Knowledge Discovery ECML-PKDD ; Conference date: 18-09-2017 Through 22-09-2017",
url = "http://ecmlpkdd2017.ijs.si/",

}

TY - CONF

T1 - Interactive Anonymization for Privacy aware Machine Learning

AU - Malle,Bernd

AU - Kieseberg,Peter

AU - Holzinger,Andreas

PY - 2017

Y1 - 2017

N2 - Privacy aware Machine Learning is the discipline of applyingMachine Learning techniques in such a way as to protect and retain personal identities during the process. This is most easily achieved by firstanonymizing a dataset before releasing it for the purpose of data mining or knowledge extraction. Starting in June 2018, this will also remainthe sole legally permitted way within the EU to release data withoutgranting people involved the right to be forgotten, i.e. the right to havetheir data deleted on request. To governments, organizations and corporations, this represents a serious impediment to research operations,since any anonymization results in a certain degree of reduced data utility. In this paper we propose applying human background knowledge viainteractive Machine Learning to the process of anonymization; this isdone by eliciting human preferences for preserving some attribute valuesover others in the light of specific tasks. Our experiments show that human knowledge can yield measurably better classification results than arigid automatic approach. However, the impact of interactive learning inthe field of anonymization will largely depend on the experimental setup,such as an appropriate choice of application domain as well as suitabletest subjects.

AB - Privacy aware Machine Learning is the discipline of applyingMachine Learning techniques in such a way as to protect and retain personal identities during the process. This is most easily achieved by firstanonymizing a dataset before releasing it for the purpose of data mining or knowledge extraction. Starting in June 2018, this will also remainthe sole legally permitted way within the EU to release data withoutgranting people involved the right to be forgotten, i.e. the right to havetheir data deleted on request. To governments, organizations and corporations, this represents a serious impediment to research operations,since any anonymization results in a certain degree of reduced data utility. In this paper we propose applying human background knowledge viainteractive Machine Learning to the process of anonymization; this isdone by eliciting human preferences for preserving some attribute valuesover others in the light of specific tasks. Our experiments show that human knowledge can yield measurably better classification results than arigid automatic approach. However, the impact of interactive learning inthe field of anonymization will largely depend on the experimental setup,such as an appropriate choice of application domain as well as suitabletest subjects.

KW - privacy aware machine learning

M3 - Paper

SP - 15

ER -