PRIVACY-PRESERVING USING BIG DATA ANALYTICS

Big in the field is a Privacy Preservation System which is a result of the integration of many data science, cryptography and privacy engineering methods which we have put together to form a strong base for doable data analytics. We see in practice that via Differential Privacy, Homomorphic Encryption, and Federated Learning we achieve privacy at every step of the analysis.    The system is put in place for privacy and security of data as well as compliance which includes global standards like GDPR and HIPAA. We have developed a web portal interface which allows our staff to easily analyze results without at the same time putting any personal data at risk.  Present study reports on the growth of privacy preserving data mining in today which we live in where we see large scale data creation in seconds. We put forth a model which is very much a success in it’s identification and handling of sensitive information in health care, banking and academic fields.  In total this research did what it set out to do which was to present a model for security, intelligence and privacy. This research also sets a base for future studies in privacy preserving computations.

1.          Adam Smith, Kobbi Nissim, Frank McSherry, and Cynthia Dwork (2006). Noise to Sensitivity Calibration in Private Data Analysis (Differential Privacy). Journal of Cryptography Theory.   2.         Shokri, R. and V. Shmatikov (2015). Deep Learning while protecting privacy. Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security (CCS).   3.         Craig Gentry (2009). Ideal lattices are used for fully homomorphic encryption. ACM Symposium on Computer Theory (STOC).   4.         McMahan, B. et al. (2017). Learning Deep Networks with Efficient Communication from Decentralized Data. The 20th International Conference on Artificial Intelligence and Statistics (AISTATS).   5.              Google AI Blog (2019). Differential Privacy: Protecting Data in the Machine Learning Era.   6.        Microsoft Research (2020). Federated Learning and Privacy-Preserving AI: Techniques and Applications. Technical Report, Microsoft Research.   7.               IBM Research (2021). Using Homomorphic Encryption to Provide Secure Data Analytics. IBM Technical Paper on Privacy Technologies.   8.          Abadi, M. et al. (2016). Deep learning's differential privacy. Proceedings of the ACM SIGSAC Conference on Computer and Communications Security.   9.            Bonawitz, K. et al. (2019). Federated Learning: Obstacles, Approaches, and Prospects. Journal of Machine Learning Systems.   10.         PySyft Documentation (2023). OpenMined Library for Privacy-Preserving Machine Learning.