An increasing number of businesses are replacing their data storage andcomputation infrastructure with cloud services. Likewise, there is an increasedemphasis on performing analytics based on multiple datasets obtained fromdifferent data sources. While ensuring security of data and computationoutsourced to a third party cloud is in itself challenging, supportinganalytics using data distributed across multiple, independent clouds is evenfurther from trivial. In this paper we present CloudMine, a cloud-based servicewhich allows multiple data owners to perform privacy-preserved computation overthe joint data using their clouds as delegates. CloudMine protects data privacywith respect to semi-honest data owners and semi-honest clouds. It furthermoreensures the privacy of the computation outputs from the curious clouds. Itallows data owners to reliably detect if their cloud delegates have been lazywhen carrying out the delegated computation. CloudMine can run as a centralizedservice on a single cloud, or as a distributed service over multiple,independent clouds. CloudMine supports a set of basic computations that can beused to construct a variety of highly complex, distributed privacy-preservingdata analytics. We demonstrate how a simple instance of CloudMine (secure sumservice) is used to implement three classical data mining tasks(classification, association rule mining and clustering) in a cloudenvironment. We experiment with a prototype of the service, the results ofwhich suggest its practicality for supporting privacy-preserving data analyticsas a (multi) cloud-based service.
展开▼
