Abstract
Federated Learning (FL), as an emerging form of distributed machine learning (ML), can protect participants’ private data from being substantially disclosed to cyber adversaries. It has potential uses in many large-scale, data-rich environments, such as the Internet of Things (IoT), Industrial IoT, Social Media (SM), and the emerging SM 3.0. However, federated learning is susceptible to some forms of data leakage through model inversion attacks. Such attacks occur through the analysis of participants’ uploaded model updates. Model inversion attacks can reveal private data and potentially undermine some critical reasons for employing federated learning paradigms. This article proposes novel differential privacy (DP)-based deep federated learning framework. We theoretically prove that our framework can fulfill DP’s requirements under distinct privacy levels by appropriately adjusting scaled variances of Gaussian noise. We then develop a Differentially Private Data-Level Perturbation (DP-DLP) mechanism to conceal any single data point’s impact on the training phase. Experiments on real-world datasets, specifically the social media 3.0, Iris, and Human Activity Recognition (HAR) datasets, demonstrate that the proposed mechanism can offer high privacy, enhanced utility, and elevated efficiency. Consequently, it simplifies the development of various DP-based FL models with different tradeoff preferences on data utility and privacy levels.
- [1] . 2020. Blockchain and federated learning for privacy-preserved data sharing in industrial IoT. IEEE Transactions on Industrial Informatics 16, 6 (2020), 4177–4186.Google Scholar
Cross Ref
- [2] . 2021. Privacy-preserving decentralized learning framework for healthcare system. ACM Transactions on Multimedia Computing, Communications, and Applications 17, 2s (2021), 1–24.Google Scholar
Digital Library
- [3] . 2021. Monitoring cyber SentiHate social behavior during COVID-19 pandemic in North America. IEEE Access 9 (2021), 91184–91208.Google Scholar
Cross Ref
- [4] . 2021. Privacy preservation in federated learning: An insightful survey from the GDPR perspective. Computers & Security 110 (2021), 102402.Google Scholar
Digital Library
- [5] . 2018. Modularity is the key a new approach to social media privacy policies. In Proceedings of the 7th Mexican Conference on Human-computer Interaction. 1–4.Google Scholar
Digital Library
- [6] 2021. 13 Critical Data Breach Stats for Australian Businesses in 2021 | UpGuard. Retrieved September 14, 2021 from https://www.upguard.com/blog/australian-data-breach-stats.Google Scholar
- [7] . 2020. Privacy-encoding models for preserving utility of machine learning algorithms in social media. In Proceedings of the 2020 IEEE 19th International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom). IEEE, 856–863.Google Scholar
Cross Ref
- [8] . 2018. Robust privacy-preserving image sharing over online social networks (OSNs). ACM Transactions on Multimedia Computing, Communications, and Applications 14, 1 (2018), 1–22.Google Scholar
Digital Library
- [9] . 2022. Social network analytic-based online counterfeit seller detection using user shared images. ACM Transactions on Multimedia Computing, Communications, and Applications (2022).Google Scholar
Digital Library
- [10] . 2021. Differentially private federated learning: Algorithm, analysis and optimization. In Proceedings of the Federated Learning Systems. Springer, 51–78.Google Scholar
Cross Ref
- [11] . 2021. Practical defences against model inversion attacks for split neural networks. arXiv:2104.05743. Retrieved from https://arxiv.org/abs/2104.05743.Google Scholar
- [12] . 2015. Model inversion attacks that exploit confidence information and basic countermeasures. In Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security. 1322–1333.Google Scholar
Digital Library
- [13] . 2020. Inverting gradients–how easy is it to break privacy in federated learning? In Advances in Neural Information Processing Systems, Vol. 33. Curran Associates, Inc., 16937–16947. https://proceedings.neurips.cc/paper/2020/file/c4ede56bbd98819ae6112b20ac6bf145-Paper.pdf.Google Scholar
- [14] . 2021. Role of machine learning and deep learning in securing 5G-driven industrial IoT applications. Ad Hoc Networks 123 (2021), 102685.Google Scholar
Digital Library
- [15] . 2019. Deep learning–based multimedia analytics: A review. ACM Transactions on Multimedia Computing, Communications, and Applications 15, 1s (2019), 1–26.Google Scholar
Digital Library
- [16] . 2020. Input perturbation: A new paradigm between central and local differential privacy. arXiv:2002.08570. Retrieved from https://arxiv.org/abs/2002.08570.Google Scholar
- [17] . 2020. Privacy-preserving traffic flow prediction: A federated learning approach. IEEE Internet of Things Journal 7, 8 (2020), 7751–7763.Google Scholar
Cross Ref
- [18] . 2021. Blockchain-enabled federated learning data protection aggregation scheme with differential privacy and homomorphic encryption in IIoT. IEEE Transactions on Industrial Informatics 18, 6 (2021), 1–1.Google Scholar
- [19] . 2016. Calibrating noise to sensitivity in private data analysis. Journal of Privacy and Confidentiality 7, 3 (2016), 17–51.Google Scholar
- [20] . 2021. A blockchain-enabled explainable federated learning for securing Internet-of-Things-based social media 3.0 networks. IEEE Transactions on Computational Social Systems (2021), 1–17.Google Scholar
Cross Ref
- [21] . 2018. Privacy-preserving social media data outsourcing. In Proceedings of the IEEE INFOCOM 2018-IEEE Conference on Computer Communications. IEEE, 1106–1114.Google Scholar
Digital Library
- [22] . 2018. Revisiting the privacy paradox: Concerns and protection strategies in the social media experiences of older adults. In Proceedings of the 9th International Conference on Social Media and Society. 150–159.Google Scholar
Digital Library
- [23] . 2018. Social media recommender systems: Review and open research issues. IEEE Access 6 (2018), 15608–15628.Google Scholar
- [24] . 2020. A trustworthy privacy preserving framework for machine learning in industrial IoT systems. IEEE Transactions on Industrial Informatics 16, 9 (2020), 6092–6102.Google Scholar
Cross Ref
- [25] . 2016. Prediction of virality timing using cascades in social media. ACM Transactions on Multimedia Computing, Communications, and Applications 13, 1 (2016), 1–23.Google Scholar
Digital Library
- [26] . 2016. Evaluating the privacy risk of user-shared images. ACM Transactions on Multimedia Computing, Communications, and Applications 12, 4s (2016), 1–21.Google Scholar
Digital Library
- [27] . 2019. A novel application recommendation method combining social relationship and trust relationship for future Internet of Things. Multimedia Tools and Applications 78, 21 (2019), 29867–29880.Google Scholar
Cross Ref
- [28] . 2019. A survey on IoT security: Application areas, security threats, and solution architectures. IEEE Access 7 (2019), 82721–82743.Google Scholar
Cross Ref
- [29] . 2021. Privacy-preserving and incentivized contact tracing for covid-19 using blockchain. IEEE Internet of Things Magazine 4, 3 (2021), 72–79.Google Scholar
Cross Ref
- [30] . 2015. De-identification of personal information. National Institute of Standards and Technology (2015), 1–46.Google Scholar
- [31] . 2018. Privacy-preserving social media data publishing for personalized ranking-based recommendation. IEEE Transactions on Knowledge and Data Engineering 31, 3 (2018), 507–520.Google Scholar
Digital Library
- [32] . 2020. Evolutionary algorithms for k-anonymity in social networks based on clustering approach. The Computer Journal 63, 7 (2020), 1039–1062.Google Scholar
Cross Ref
- [33] . 2017. Privacy-preserving data mining: Methods, metrics, and applications. IEEE Access 5 (2017), 10562–10582.Google Scholar
Cross Ref
- [34] . 2021. Anonymization techniques for privacy preserving data publishing: A comprehensive survey. IEEE Access 9 (2021), 8512–8545.Google Scholar
- [35] . 2017. Differentially private federated learning: A client level perspective. arXiv:1712.07557. Retrieved from https://arxiv.org/abs/1712.07557.Google Scholar
- [36] . 2020. Federated learning and differential privacy: Software tools analysis, the Sherpa. ai FL framework and methodological guidelines for preserving data privacy. Information Fusion 64 (2020), 270–292.Google Scholar
Cross Ref
- [37] . 2019. A hybrid approach to privacy-preserving federated learning. In Proceedings of the 12th ACM Workshop on Artificial Intelligence and Security. 1–11.Google Scholar
Digital Library
- [38] . 2022. Data analytics of social media 3.0: Privacy protection perspectives for integrating social media and Internet of Things (SM-IoT) systems. Ad Hoc Networks 128 (2022), 102786.
DOI: Google ScholarDigital Library
- [39] . 2017. UCI Machine Learning Repository. Retrieved Accessed 27 January, 2022 from http://archive.ics.uci.edu/ml.Google Scholar
- [40] . 2013. A Public Domain Dataset for Human Activity Recognition Using Smartphones. Retrieved January 27, 2022 from https://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones.Google Scholar
- [41] . 2018. Blockchain-based privacy preserving deep learning. In Proceedings of the International Conference on Information Security and Cryptology. Springer, 370–383.Google Scholar
- [42] . 2020. Deep learning models for real-time human activity recognition with smartphones. Mobile Networks and Applications 25, 2 (2020), 743–755.Google Scholar
Digital Library
Index Terms
Perturbation-enabled Deep Federated Learning for Preserving Internet of Things-based Social Networks
Recommendations
Privacy preserving data obfuscation for inherently clustered data
Privacy is defined as the freedom from unauthorised intrusion. The availability of public records along with intelligent search engines and data mining tools allow easy access to useful information. They also serve as a haven for individuals with ...
A review of privacy preserving models for multi-party data release framework
WIR '16: Proceedings of the ACM Symposium on Women in Research 2016Nowadays, with the improvement of internet technology and advancement in distributed computing data is increasing rapidly. There is a need of information sharing between organizations. Ideally, we wish to share data from multiple private databases and ...
Privacy preserving techniques for decision trees
As a representative classification model, decision tree has been extensively applied in data mining. It generates a series of if-then rules based on the homogeneity of class distribution. In a society where data spreads everywhere for knowledge discovery, ...






Comments