Abstract
Preserving privacy in contemporary NLP models allows us to work with sensitive data, but unfortunately comes at a price. We know that stricter privacy guarantees in differentially-private stochastic gradient descent (DP-SGD) generally degrade model performance. However, previous research on the efficiency of DP-SGD in NLP is inconclusive or even counter-intuitive. In this short paper, we provide an extensive analysis of different privacy preserving strategies on seven downstream datasets in five different ‘typical’ NLP tasks with varying complexity using modern neural models based on BERT and XtremeDistil architectures. We show that unlike standard non-private approaches to solving NLP tasks, where bigger is usually better, privacy-preserving strategies do not exhibit a winning pattern, and each task and privacy regime requires a special treatment to achieve adequate performance.- Anthology ID:
- 2022.emnlp-main.496
- Volume:
- Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
- Month:
- December
- Year:
- 2022
- Address:
- Abu Dhabi, United Arab Emirates
- Editors:
- Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 7340–7353
- Language:
- URL:
- https://aclanthology.org/2022.emnlp-main.496
- DOI:
- 10.18653/v1/2022.emnlp-main.496
- Cite (ACL):
- Manuel Senge, Timour Igamberdiev, and Ivan Habernal. 2022. One size does not fit all: Investigating strategies for differentially-private learning across NLP tasks. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 7340–7353, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
- Cite (Informal):
- One size does not fit all: Investigating strategies for differentially-private learning across NLP tasks (Senge et al., EMNLP 2022)
- PDF:
- https://preview.aclanthology.org/naacl24-info/2022.emnlp-main.496.pdf