Abstract
With federated learning, information among different clients can be accessed to train a central model that aims for an optimal use of data while keeping the clients’ data local and private. But since its emergence in 2017, several threats such as gradient attacks or model poisoning attacks against federated learning have been identified. Therefore, federated learning cannot be considered as stand alone privacy preserving machine learning technique. Thus, we analyse how and where local differential privacy can compensate for the drawbacks of federated learning while keeping its advantage of combining data from different sources. In this work, we analyse the different communication channels and entities in the federated learning architecture that may be attacked or try to reveal data from other entities. Thereby, we evaluate where local differential privacy is helpful. Finally, for our spam and ham email classification model with local differential privacy, we find that setting a local threshold of F1-Score on the clients’ level can reduce the consumption of privacy budget over several rounds, and decrease the training time. Moreover, we find that for the central model a significantly higher F1-Score than those set on the local level for the clients can be achieved.
This work was supported by the European Union’s Horizon 2020 Research and Innovation Program through the Project CyberSec4Europe under Agreement 830929.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Abadi, M., et al.: Deep learning with differential privacy. In: Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, pp. 308–318 (2016)
Amjad, M., Voronkov, I., Saenko, A., Gelbukh, A.: Comparison of text classification methods using deep learning neural networks. In: Proceedings of the 20th International Conference on Computational Linguistics and Intelligent Text Processing (CICLing) (2019)
Awan, S., Luo, B., Li, F.: CONTRA: defending against poisoning attacks in federated learning. In: Bertino, E., Shulman, H., Waidner, M. (eds.) ESORICS 2021. LNCS, vol. 12972, pp. 455–475. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-88418-5_22
Basu, P., Roy, T.S., Naidu, R., Muftuoglu, Z.: Privacy enabled financial text classification using differential privacy and federated learning. arXiv preprint arXiv:2110.01643 (2021)
Bhowmick, A., Hazarika, S.M.: E-mail spam filtering: a review of techniques and trends. In: Kalam, A., Das, S., Sharma, K. (eds.) Advances in Electronics, Communication and Computing. LNEE, vol. 443, pp. 583–590. Springer, Singapore (2018). https://doi.org/10.1007/978-981-10-4765-7_61
Dada, E.G., Bassi, J.S., Chiroma, H., Adetunmbi, A.O., Ajibuwa, O.E., et al.: Machine learning for email spam filtering: review, approaches and open research problems. Heliyon 5(6), e01802 (2019)
Dong, Y., Chen, X., Li, K., Wang, D., Zeng, S.: FLOD: oblivious defender for private Byzantine-robust federated learning with dishonest-majority. Cryptology ePrint Archive (2021)
Dwork, C., McSherry, F., Nissim, K., Smith, A.: Calibrating noise to sensitivity in private data analysis. In: Halevi, S., Rabin, T. (eds.) TCC 2006. LNCS, vol. 3876, pp. 265–284. Springer, Heidelberg (2006). https://doi.org/10.1007/11681878_14
Geiping, J., Bauermeister, H., Dröge, H., Moeller, M.: Inverting gradients-how easy is it to break privacy in federated learning? In: Advances in Neural Information Processing Systems, vol. 33, pp. 16937–16947 (2020)
Geyer, R.C., Klein, T., Nabi, M.: Differentially private federated learning: a client level perspective. arXiv preprint arXiv:1712.07557 (2017)
Jain, G., Sharma, M., Agarwal, B.: Optimizing semantic LSTM for spam detection. Int. J. Inf. Technol. 11(2) (2019)
Li, T., Sahu, A.K., Talwalkar, A., Smith, V.: Federated learning: challenges, methods, and future directions. IEEE Signal Process. Mag. 37(3), 50–60 (2020)
Löbner, S., Tesfay, W.B., Nakamura, T., Pape, S.: Explainable machine learning for default privacy setting prediction. IEEE Access 9, 63700–63717 (2021)
Löbner, S., Tronnier, F., Pape, S., Rannenberg, K.: Comparison of de-identification techniques for privacy preserving data analysis in vehicular data sharing. In: Computer Science in Cars Symposium, pp. 1–11 (2021)
Makkar, A., Ghosh, U., Rawat, D.B., Abawajy, J.: FedLearnSP: preserving privacy and security using federated learning and edge computing. IEEE Consum. Electron. Mag. 11, 21–27 (2021)
McMahan, B., Moore, E., Ramage, D., Hampson, S., Arcas, B.A.: Communication-efficient learning of deep networks from decentralized data. In: Artificial Intelligence and Statistics, pp. 1273–1282. PMLR (2017)
Metsis, V., Androutsopoulos, I., Paliouras, G.: Spam filtering with Naive Bayes-which Naive Bayes? In: CEAS, Mountain View, CA, vol. 17 (2006)
Paszke, A., et al.: PyTorch: an imperative style, high-performance deep learning library. In: Wallach, H., Larochelle, H., Beygelzimer, A., d’Alché-Buc, F., Fox, E., Garnett, R. (eds.) Advances in Neural Information Processing Systems, vol. 32, pp. 8024–8035. Curran Associates, Inc. (2019)
Powers, D.M.: Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation. preprint arXiv:2010.16061 (2020)
Tanuwidjaja, H.C., Choi, R., Baek, S., Kim, K.: Privacy-preserving deep learning on machine learning as a service-a comprehensive survey. IEEE Access 8, 167425–167447 (2020)
Thapa, C., et al.: FedEmail: performance measurement of privacy-friendly phishing detection enabled by federated learning. arXiv - CS - Machine Learning (2020)
Tolpegin, V., Truex, S., Gursoy, M.E., Liu, L.: Data poisoning attacks against federated learning systems. In: Chen, L., Li, N., Liang, K., Schneider, S. (eds.) ESORICS 2020. LNCS, vol. 12308, pp. 480–501. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58951-6_24
Triastcyn, A., Faltings, B.: Federated learning with Bayesian differential privacy. In: 2019 IEEE International Conference on Big Data (Big Data), pp. 2587–2596. IEEE (2019)
Tronnier, F., Pape, S., Löbner, S., Rannenberg, K.: A discussion on ethical cybersecurity issues in digital service chains. In: Kołodziej, J., Repetto, M., Duzha, A. (eds.) Cybersecurity of Digital Service Chains. LNCS, vol. 13300, pp. 222–256. Springer, Cham (2022). https://doi.org/10.1007/978-3-031-04036-8_10
Wei, K., et al.: Federated learning with differential privacy: algorithms and performance analysis. IEEE Trans. Inf. Forensics Secur. 15, 3454–3469 (2020)
Wei, W., et al.: A framework for evaluating client privacy leakages in federated learning. In: Chen, L., Li, N., Liang, K., Schneider, S. (eds.) ESORICS 2020. LNCS, vol. 12308, pp. 545–566. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58951-6_27
Yang, Q., Liu, Y., Chen, T., Tong, Y.: Federated machine learning: concept and applications. ACM Trans. Intell. Syst. Technol. (TIST) 10(2), 1–19 (2019)
Yousefpour, A., et al.: Opacus: user-friendly differential privacy library in PyTorch. arXiv preprint arXiv:2109.12298 (2021)
Zhu, L., Liu, Z., Han, S.: Deep leakage from gradients. In: Advances in Neural Information Processing Systems, vol. 32 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Löbner, S., Gogov, B., Tesfay, W.B. (2023). Enhancing Privacy in Federated Learning with Local Differential Privacy for Email Classification. In: Garcia-Alfaro, J., Navarro-Arribas, G., Dragoni, N. (eds) Data Privacy Management, Cryptocurrencies and Blockchain Technology. DPM CBT 2022 2022. Lecture Notes in Computer Science, vol 13619. Springer, Cham. https://doi.org/10.1007/978-3-031-25734-6_1
Download citation
DOI: https://doi.org/10.1007/978-3-031-25734-6_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-25733-9
Online ISBN: 978-3-031-25734-6
eBook Packages: Computer ScienceComputer Science (R0)