Implicit privacy preservation: a framework based on data generation
This paper addresses a special and imperceptible class of privacy, called implicit privacy. In contrast to traditional (explicit) privacy, implicit privacy has two essential properties: (1) It is not initially defined as a privacy attribute; (2) it is strongly associated with privacy attributes. In...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2022-01-01
|
| Series: | Security and Safety |
| Subjects: | |
| Online Access: | https://sands.edpsciences.org/articles/sands/full_html/2022/01/sands20220006/sands20220006.html |
| _version_ | 1797430461528539136 |
|---|---|
| author | Yang Qing Wang Cheng Hu Teng Chen Xue Jiang Changjun |
| author_facet | Yang Qing Wang Cheng Hu Teng Chen Xue Jiang Changjun |
| author_sort | Yang Qing |
| collection | DOAJ |
| description | This paper addresses a special and imperceptible class of privacy, called implicit privacy. In contrast to traditional (explicit) privacy, implicit privacy has two essential properties: (1) It is not initially defined as a privacy attribute; (2) it is strongly associated with privacy attributes. In other words, attackers could utilize it to infer privacy attributes with a certain probability, indirectly resulting in the disclosure of private information. To deal with the implicit privacy disclosure problem, we give a measurable definition of implicit privacy, and propose an ex-ante implicit privacy-preserving framework based on data generation, called IMPOSTER. The framework consists of an implicit privacy detection module and an implicit privacy protection module. The former uses normalized mutual information to detect implicit privacy attributes that are strongly related to traditional privacy attributes. Based on the idea of data generation, the latter equips the Generative Adversarial Network (GAN) framework with an additional discriminator, which is used to eliminate the association between traditional privacy attributes and implicit ones. We elaborate a theoretical analysis for the convergence of the framework. Experiments demonstrate that with the learned generator, IMPOSTER can alleviate the disclosure of implicit privacy while maintaining good data utility. |
| first_indexed | 2024-03-09T09:27:55Z |
| format | Article |
| id | doaj.art-54ce8375767b49b3b79c43661ff70ab8 |
| institution | Directory Open Access Journal |
| issn | 2826-1275 |
| language | English |
| last_indexed | 2024-03-09T09:27:55Z |
| publishDate | 2022-01-01 |
| publisher | EDP Sciences |
| record_format | Article |
| series | Security and Safety |
| spelling | doaj.art-54ce8375767b49b3b79c43661ff70ab82023-12-02T05:13:31ZengEDP SciencesSecurity and Safety2826-12752022-01-011202200810.1051/sands/2022008sands20220006Implicit privacy preservation: a framework based on data generationYang Qing0https://orcid.org/0000-0002-5092-5437Wang Cheng1https://orcid.org/0000-0002-4752-0316Hu Teng2https://orcid.org/0000-0003-4946-8977Chen Xue3https://orcid.org/0000-0002-0391-9376Jiang Changjun4https://orcid.org/0000-0003-0637-9317Key Laboratory of Embedded System and Service Computing (Tongji University), Ministry of EducationKey Laboratory of Embedded System and Service Computing (Tongji University), Ministry of EducationKey Laboratory of Embedded System and Service Computing (Tongji University), Ministry of EducationKey Laboratory of Embedded System and Service Computing (Tongji University), Ministry of EducationKey Laboratory of Embedded System and Service Computing (Tongji University), Ministry of EducationThis paper addresses a special and imperceptible class of privacy, called implicit privacy. In contrast to traditional (explicit) privacy, implicit privacy has two essential properties: (1) It is not initially defined as a privacy attribute; (2) it is strongly associated with privacy attributes. In other words, attackers could utilize it to infer privacy attributes with a certain probability, indirectly resulting in the disclosure of private information. To deal with the implicit privacy disclosure problem, we give a measurable definition of implicit privacy, and propose an ex-ante implicit privacy-preserving framework based on data generation, called IMPOSTER. The framework consists of an implicit privacy detection module and an implicit privacy protection module. The former uses normalized mutual information to detect implicit privacy attributes that are strongly related to traditional privacy attributes. Based on the idea of data generation, the latter equips the Generative Adversarial Network (GAN) framework with an additional discriminator, which is used to eliminate the association between traditional privacy attributes and implicit ones. We elaborate a theoretical analysis for the convergence of the framework. Experiments demonstrate that with the learned generator, IMPOSTER can alleviate the disclosure of implicit privacy while maintaining good data utility.https://sands.edpsciences.org/articles/sands/full_html/2022/01/sands20220006/sands20220006.htmlprivacy preservationimplicit privacygenerative adversarial networkdata utilitydata generation |
| spellingShingle | Yang Qing Wang Cheng Hu Teng Chen Xue Jiang Changjun Implicit privacy preservation: a framework based on data generation Security and Safety privacy preservation implicit privacy generative adversarial network data utility data generation |
| title | Implicit privacy preservation: a framework based on data generation |
| title_full | Implicit privacy preservation: a framework based on data generation |
| title_fullStr | Implicit privacy preservation: a framework based on data generation |
| title_full_unstemmed | Implicit privacy preservation: a framework based on data generation |
| title_short | Implicit privacy preservation: a framework based on data generation |
| title_sort | implicit privacy preservation a framework based on data generation |
| topic | privacy preservation implicit privacy generative adversarial network data utility data generation |
| url | https://sands.edpsciences.org/articles/sands/full_html/2022/01/sands20220006/sands20220006.html |
| work_keys_str_mv | AT yangqing implicitprivacypreservationaframeworkbasedondatageneration AT wangcheng implicitprivacypreservationaframeworkbasedondatageneration AT huteng implicitprivacypreservationaframeworkbasedondatageneration AT chenxue implicitprivacypreservationaframeworkbasedondatageneration AT jiangchangjun implicitprivacypreservationaframeworkbasedondatageneration |