Models and Algorithms for Private Data Sharing

摘要: In recent years, there has been a tremendous growth in the collection of digital information about individuals. Many organizations such as governmental agencies, hospitals, and financial companies collect disseminate various person-specific data. Due to rapid advance storing, processing, networking capabilities computing devices, collected data can now be easily analyzed infer valuable for research business purposes. Data from different sources integrated further gain better insights. On one hand, offer opportunities mining useful information. other process poses threat individual privacy since these often contain sensitive this thesis, we address problem developing anonymization algorithms thwart potential attacks real-life sharing scenarios. particular, study two models: LKC-privacy differential privacy. For each models, develop anonymizing types relational data, trajectory heterogeneous We also distributed where multiple publishers cooperate integrate their private without violating given privacy requirements. Experimental results on demonstrate that proposed effectively retain essential analysis are scalable large sets.