Outsourcing collaboration analysis of multiparty privacy data using the improved Yannakakis
摘要
Many organizations are producing or collecting private data in fields such as medical research and government regulation. Due to current privacy protection, laws and regulations, commercial competition, and other issues, these institutions cannot directly share their data. Collaborative analysis of private data from multiple institutions will benefit each institution and create profits together. Therefore, we propose a Yannakakis-based multiparty outsourcing collaboration analysis scheme. It enables organizations to collaboratively analyze private data from multiple organizations according to their needs while ensuring that private data are not leaked to each other. Our scheme is based on the improved Yannakakis algorithm to build a series of query components, such as Semi-join, Join, Order-by, etc. We also optimized the join operation. By confusing the input tuples and protecting their authenticity through annotations, the join operation can be directly joined through the hash value without disclosing the join results. Through this series of configurations, we can execute a query with