A Differential Privacy-based Secure Data Sharing Model in Cloud Environment

With the large variety of cloud services, including on-demand service, least cost, and maximum scalability, it also provides machine learning services that are utilized in various applications, such as medical diagnosis, risk analysis, and image recognition. Due to numerous benefits, most organizations transfer their data to the cloud platform for storage, analysis, and sharing purposes. As the cloud is run by a third party; therefore, data owners cannot fully rely on this environment. Thus, maintaining owners' data protection while sharing it with other parties has become a major challenge. This paper proposes a secure data sharing model in the cloud environment, which protects the owners' data by encryption, noise addition, and performing the classification task by employing access control, differential privacy, and machine learning mechanisms, respectively. The proposed model's performance (in terms of accuracy, precision, recall, and F1-score) is assessed over Dermatology, Ecoli, Yeast, and Cryotherapy datasets. The experimental results exhibit that the proposed model has high accuracy, precision, recall, and F1-score up to 97.29%, 97.50%, 96.52%, and 97.25%, respectively.