Differentially Private Instance Encoding against Privacy Attacks

Shangyu Xie, Yuan Hong


Abstract
TextHide was recently proposed to protect the training data via instance encoding in natural language domain. Due to the lack of theoretic privacy guarantee, such instance encoding scheme has been shown to be vulnerable against privacy attacks, e.g., reconstruction attack. To address such limitation, we revise the instance encoding scheme with differential privacy and thus provide a provable guarantee against privacy attacks. The experimental results also show that the proposed scheme can defend against privacy attacks while ensuring learning utility (as a trade-off).
Anthology ID:
2022.naacl-srw.22
Volume:
Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Student Research Workshop
Month:
July
Year:
2022
Address:
Hybrid: Seattle, Washington + Online
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
172–180
Language:
URL:
https://aclanthology.org/2022.naacl-srw.22
DOI:
10.18653/v1/2022.naacl-srw.22
Bibkey:
Cite (ACL):
Shangyu Xie and Yuan Hong. 2022. Differentially Private Instance Encoding against Privacy Attacks. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Student Research Workshop, pages 172–180, Hybrid: Seattle, Washington + Online. Association for Computational Linguistics.
Cite (Informal):
Differentially Private Instance Encoding against Privacy Attacks (Xie & Hong, NAACL 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/remove-xml-comments/2022.naacl-srw.22.pdf
Video:
 https://preview.aclanthology.org/remove-xml-comments/2022.naacl-srw.22.mp4
Data
CoLASST