Adaptive Feature Representation Learning for Privacy-Fairness Joint Optimization
Source abstract
Abstract
Coded text representations often contain a large amount of personal sensitive information, which can easily lead to problems such as privacy leakage and model prediction bias. Most of the existing methods focus on optimizing a single objective, making it difficult to achieve an effective balance between model performance, fairness and privacy protection. For this reason, this paper proposes a new adaptive feature representation learning method, AMF-DP (adaptive matrix factorization with differential privacy). The method combines adaptive matrix factorization with a differential privacy technique to effectively improve the fairness of the model while realizing privacy protection. The experimental results show that AMF-DP is able to achieve a better balance between privacy protection, fairness, and model performance, providing a new way of thinking for text feature representation learning that takes into account multi-objective optimization.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Privacy-Preserving Technologies in Data · Privacy, Security, and Data Protection · Advanced Graph Neural Networks
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China
Electric Power Research Institute, State Grid Heilongjiang Electric Power Co., Ltd., Harbin 150040, China
Funding
National Natural Science Foundation of China · Key Research and Development Program of Heilongjiang
References (44)
Show source-provided references
1. Li, H., Xu, M., and Song, Y. (2023). Sentence embedding leaks more information than you expect: Generative embedding inversion attack to recover the whole sentence. arXiv.10.18653/v1/2023.findings-acl.881 ↗
2. Lehman, E., Jain, S., Pichotta, K., Goldberg, Y., and Wallace, B.C. (2021). Does BERT pretrained on clinical notes reveal sensitive data?. arXiv.10.18653/v1/2021.naacl-main.73 ↗
3. Pan, X., Zhang, M., Ji, S., and Yang, M. (2020, January 18–20). Privacy risks of general-purpose language models. Proceedings of the IEEE Symposium on Security and Privacy (SP), San Francisco, CA, USA.10.1109/SP40000.2020.00095 ↗
4. Sun, T., Gaut, A., Tang, S., Huang, Y., Elsherief, M., Zhao, J., Mirza, D., Belding-Royer, E.M., Chang, K.-W., and Wang, W.Y. (2019). Mitigating Gender Bias in Natural Language Processing: Literature Review. arXiv.10.18653/v1/P19-1159 ↗
5. Fairness and Bias in Algorithmic Hiring: A Multidisciplinary Survey10.1145/3696457 ↗
6. Cross, J.L., Choma, M.A., and Onofrey, J.A. (2024). Bias in medical AI: Implications for clinical decision-making. PLoS Digit. Health, 3.10.1371/journal.pdig.0000651 ↗
7. Phan, N., Jin, R., Thai, M.T., Hu, H., and Dou, D. (2019). Preserving Differential Privacy in Adversarial Learning with Provable Robustness. arXiv.10.24963/ijcai.2019/660 ↗
8. Panda, A., Wu, T., Wang, J.T., and Mittal, P. (2023, January 1–5). Privacy-Preserving In-Context Learning for Large Language Models. Proceedings of the International Conference on Learning Representations (ICLR), Kigali, Rwanda.
9. Bagdasarian, E., and Shmatikov, V. (2019). Differential Privacy Has Disparate Impact on Model Accuracy. arXiv.
10. Yang, M., Ding, M., Qu, Y., Ni, W., Smith, D., and Rakotoarivelo, T. (2024). Privacy at a Price: Exploring its Dual Impact on AI Fairness. arXiv.
11. Zhang, B., Lemoine, B., and Mitchell, M. (2018, January 2–3). Mitigating Unwanted Biases with Adversarial Learning. Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society, New Orleans, LA, USA.10.1145/3278721.3278779 ↗
12. Roh, Y., Lee, K., Whang, S.E., and Suh, C. (2020). FairBatch: Batch Selection for Model Fairness. arXiv.
13. Lahoti, P., Beutel, A., Chen, J., Lee, K., Prost, F., Thain, N., Wang, X., and Chi, E.H. (2020). Fairness without Demographics through Adversarially Reweighted Learning. arXiv.
14. Towards Fair and Privacy-Preserving Federated Deep Models10.1109/TPDS.2020.2996273 ↗
15. Yang, Z., Ge, Y., Su, C., Wang, D., Zhao, X., and Ying, Y. (May, January 30). Fairness-aware Differentially Private Collaborative Filtering. Proceedings of the WWW ’23 Companion: Companion Proceedings of the ACM Web Conference 2023, Austin, TX, USA.10.1145/3543873.3587577 ↗
16. Ghoukasian, H., and Asoodeh, S. (2024, January 7–12). Differentially Private Fair Binary Classifications. Proceedings of the 2024 IEEE International Symposium on Information Theory (ISIT), Athens, Greece.10.1109/ISIT57864.2024.10619147 ↗
17. Staab, R., Vero, M., Balunović, M., and Vechev, M. (2023). Beyond Memorization: Violating Privacy via Inference with Large Language Models. arXiv.
18. Ghazi, B., Huang, Y., Kamath, P., Kumar, R., Manurangsi, P., Sinha, A., and Zhang, C. (2023). Sparsity-Preserving Differentially Private Training of Large Embedding Models. arXiv.
19. Smith, V., Shamsabadi, A.S., Ashurst, C., and Weller, A. (2023). Identifying and Mitigating Privacy Risks Stemming from Language Models: A Survey. arXiv.
20. Differentially Private Recurrent Variational Autoencoder for Text Privacy Preservation10.1007/s11036-023-02096-9 ↗
21. Silva, B.C.C.D., Ferraz, T.P., and Lopes, R.D.D. (2024). Enriching GNNs with Text Contextual Representations for Detecting Disinformation Campaigns on Social Media. arXiv.
22. Privacy and Fairness in Federated Learning: On the Perspective of Tradeoff
23. Niu, B., Chen, Y., Wang, B., Cao, J., and Li, F. (2020, January 25–28). Utility-Aware Exponential Mechanism for Personalized Differential Privacy. Proceedings of the 2020 IEEE Wireless Communications and Networking Conference (WCNC), Seoul, Republic of Korea.10.1109/WCNC45663.2020.9120532 ↗
24. Blodgett, S.L., Barocas, S., Daumé, H., and Wallach, H.M. (2020). Language (Technology) is Power: A Critical Survey of “Bias” in NLP. arXiv.10.18653/v1/2020.acl-main.485 ↗
25. Ravfogel, S., Elazar, Y., Gonen, H., Twiton, M., and Goldberg, Y. (2020, January 5–10). Null It Out: Guarding Protected Attributes by Iterative Nullspace Projection. Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL), Seattle, WA, USA.10.18653/v1/2020.acl-main.647 ↗
26. Sun, C., Xu, C., Yao, C., Liang, S., Wu, Y., Liang, D., Liu, X., and Liu, A. (2022). Improving Robust Fairness via Balance Adversarial Training. arXiv.10.1609/aaai.v37i12.26769 ↗
27. Yao, W., Zhou, Z., Li, Z., Han, B., and Liu, Y. (2023). Understanding Fairness Surrogate Functions in Algorithmic Fairness. arXiv.
28. Yang, J., Jiang, J., Sun, Z., and Chen, J. (2024, January 15–19). A Large-Scale Empirical Study on Improving the Fairness of Image Classification Models. Proceedings of the 33rd ACM SIGSOFT International Symposium on Software Testing and Analysis, Vienna, Austria.10.1145/3650212.3652122 ↗
29. The Effect of Resampling Techniques on the Performances of Machine Learning Clinical Risk Prediction Models in the Setting of Severe Class Imbalance: Development and Internal Validation in a Retrospective Cohort10.1007/s44163-024-00199-0 ↗
30. Lin, Y., Li, D., Zhao, C., Wu, X., Tian, Q., and Shao, M. (2024, January 3–9). Supervised Algorithmic Fairness in Distribution Shifts: A Survey. Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI), Jeju, Republic of Korea.
31. Palakkadavath, R., Le, H., Nguyen-Tang, T., Gupta, S., and Venkatesh, S. (March, January 26). Fair Domain Generalization with Heterogeneous Sensitive Attributes Across Domains. Proceedings of the 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Tucson, AZ, USA.10.1109/WACV61041.2025.00718 ↗
32. Maheshwari, G., Denis, P., Keller, M., and Bellet, A. (2022). Fair NLP Models with Differentially Private Text Encoders. arXiv.10.18653/v1/2022.findings-emnlp.514 ↗
33. Wallach, H., Larochelle, H., Beygelzimer, A., d’Alché-Buc, F., Fox, E., and Garnett, R. (2019). Differential Privacy Has Disparate Impact on Model Accuracy. Advances in Neural Information Processing Systems 32 (NeurIPS 2019), Curran Associates, Inc.
34. Matrix Factorization Techniques for Recommender Systems10.1109/MC.2009.263 ↗
35. Liang, D., Krishnan, R.G., Hoffman, M.D., and Jebara, T. (2018, January 23–27). Variational Autoencoders for Collaborative Filtering. Proceedings of the 2018 World Wide Web Conference (WWW), Lyon, France.10.1145/3178876.3186150 ↗
36. Sedhain, S., Menon, A.K., Sanner, S., and Xie, L. (2015, January 18–22). AutoRec: Autoencoders Meet Collaborative Filtering. Proceedings of the 24th International Conference on World Wide Web (WWW), Florence, Italy.10.1145/2740908.2742726 ↗
37. Blodgett, S.L., Green, L., and O’Connor, B. (2016). Demographic Dialectal Variation in Social Media: A Case Study of African-American English. arXiv.10.18653/v1/D16-1120 ↗
38. De-Arteaga, M., Romanov, A., Wallach, H., Chayes, J., Borgs, C., Chouldechova, A., Geyik, S., Kenthapadi, K., and Kalai, A.T. (2019, January 29–31). Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting. Proceedings of the Conference on Fairness, Accountability, and Transparency (FAccT), Atlanta, GA, USA.10.1145/3287560.3287572 ↗
39. Kohavi, R. (1996, January 2–4). Scaling Up the Accuracy of Naive-Bayes Classifiers: A Decision-Tree Hybrid. Proceedings of the Second International Conference on Knowledge Discovery and Data Mining (KDD), Portland, OR, USA.
40. Romanov, A., De-Arteaga, M., Wallach, H., Chayes, J., Borgs, C., Chouldechova, A., Geyik, S., Kenthapadi, K., Rumshisky, A., and Kalai, A.T. (2019). What’s in a Name? Reducing Bias in Bios without Access to Protected Attributes. arXiv.10.1145/3287560.3287572 ↗
41. Voita, E., and Titov, I. (2020). Information-Theoretic Probing with Minimum Description Length. arXiv.10.18653/v1/2020.emnlp-main.14 ↗
42. Xu, D., Yuan, S., Zhang, L., and Wu, X. (2018, January 10–13). FairGAN: Fairness-Aware Generative Adversarial Networks. Proceedings of the 2018 IEEE International Conference on Big Data (Big Data), Seattle, WA, USA.10.1109/BigData.2018.8622525 ↗
43. Abadi, M., Chu, A., Goodfellow, I., McMahan, H.B., Mironov, I., Talwar, K., and Zhang, L. (2016, January 24–28). Deep Learning with Differential Privacy. Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security (CCS), Vienna, Austria.10.1145/2976749.2978318 ↗
44. Hua, J., Xia, C., and Zhong, S. (2015, January 25–31). Differentially Private Matrix Factorization. Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI), Buenos Aires, Argentina.