Privacy-preserving Active Learning on Sensitive Data for User Intent Classification

Oluwaseyi Feyisetan, Thomas Drake, Borja Balle, Tom Diethe

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

21 Downloads (Pure)

Abstract

Active learning holds promise of significantly reducing data annotation costs while maintaining reasonable model performance. However, it requires sending data to annotators for labeling. This presents a possible privacy leak when the training set includes sensitive user data. In this paper, we describe an approach for carrying out privacy preserving active learning with quantifiable guarantees. We evaluate our approach by showing the tradeoff between privacy, utility and annotation budget on a binary classification task in a active learning setting.
Original languageEnglish
Title of host publicationProceedings of the PAL: Privacy-Enhancing Artificial Intelligence and Language Technologies
Subtitle of host publicationAs Part of the AAAI Spring Symposium Series (AAAI-SSS 2019)
PublisherCEUR Workshop Proceedings
Publication statusPublished - 26 Mar 2019

Publication series

NameCEUR Workshop Proceedings
Volume2335
ISSN (Print)1613-0073

Bibliographical note

To appear at PAL: Privacy-Enhancing Artificial Intelligence and Language Technologies as part of the AAAI Spring Symposium Series (AAAI-SSS 2019)

Keywords

  • cs.LG
  • cs.CL
  • stat.ML

Fingerprint

Dive into the research topics of 'Privacy-preserving Active Learning on Sensitive Data for User Intent Classification'. Together they form a unique fingerprint.

Cite this