| The rapidly increasing prevalence of distributed data-driven applications highlights security and privacy issues in storing and processing sensitive data. Although manipulating raw data may violate privacy of their owners, techniques for processing and using privacy-preserving data descriptions can help. It remains a challenge, however, to ensure that adapted and new solutions are efficient, secure, and preserve privacy of data owners without disclosing confidentiality of data providers.;This dissertation proposes a new notion of m-privacy that addresses situations in which data providers may act as adversaries. To verify if such adversaries are capable of breaching privacy, we introduce novel strategies and an adaptive algorithm to select and use the most efficient approach. In addition, we design an algorithm to anonymize data to be m-private, i.e., any m colluding parties cannot compromise privacy. All verification and anonymization algorithms are implemented to be run in distributed environments by a trusted third party.;For settings without a trusted third party, we introduce new secure multiparty computation protocols that implement m-privacy verification and anonymization algorithms. For each protocol, we prove its security, analyze its communication complexity, and evaluate its overall performance for various settings.;This dissertation also describes a new two-phase algorithm to release differentially private histograms for records with customized privacy levels. We adapt a v-optimal partitioning algorithm to make it usable with differential privacy, and experimentally evaluate its performance.;Finally, for settings without a trusted third party, this dissertation presents a new distributed differential privacy mechanism that achieves collusion resistance with small overhead. We also define an enhanced fault tolerant and secure scheme for multiparty aggregation operations, and we employ it to implement our differential privacy mechanism in distributed environments. Both the privacy mechanism and the fault tolerant scheme are extensively analyzed and experimentally evaluated. |
