2019/03/31 by M. Suresh, Suresh, Madhav, Zuohao She +8
Computer Science · #Blockchain Technology Applications and Security #Cloud Data Security Solutions #Cryptography and Data Security #Cryptography and Security (cs.CR) #Databases (cs.DB) #FOS: Computer and information sciences #Privacy-Preserving Technologies in Data
paper · pdf · doi:10.48550/arxiv.1904.00411
openalex publication_date 2019/03/31 · openalex created_date 2022/07/29 · openalex updated_date 2026/07/28
A private data federation enables data owners to pool their information for\nquerying without disclosing their secret tuples to one another. Here, a client\nqueries the union of the records of all data owners. The data owners work\ntogether to answer the query using privacy-preserving algorithms that prevent\nthem from learning unauthorized information about the inputs of their peers.\nOnly the client, and a federation coordinator, learn the query's output.\nKloakDB is a private data federation that uses trusted hardware to process SQL\nqueries over the inputs of two or more parties. Currently private data\nfederations compute their queries fully-obliviously, guaranteeing that no\ninformation is revealed about the sensitive inputs of a data owner to their\npeers by observing the query's instruction traces and memory access patterns.\nOblivious querying almost always exacts multiple orders of magnitude slowdown\nin query runtimes compared to plaintext execution, making it impractical for\nmany applications. KloakDB offers a semi-oblivious computing framework,\nk-anonymous query processing. We make the query's observable transcript\nk-anonymous because it is a popular standard for data release in many domains\nincluding medicine, educational research, and government data. KloakDB's\nqueries run such that each data owner may deduce information about no fewer\nthan k individuals in the data of their peers. In addition, stakeholders set\nk, creating a novel trade-off between privacy and performance. Our results\nshow that KloakDB enjoys speedups of up to 117X using k-anonymous query\nprocessing over full-oblivious evaluation.\n