ABSTRACT:
Today’s organizations raise an expanding requirement for information sharing by means of on-demand access. Information brokering systems (IBS) have been proposed to associate large scale loosely federated data sources via a brokering overlay, in which the representatives settle on routing decisions to guide customer queries to the requesting data servers. Many existing IBSs expect that specialists are trusted and therefore just receive server-side access control for information privacy. In any case, the privacy of data location and data consumer can even now be derived from metadata, (for example, question and access control rules) traded inside the IBS, yet little consideration has been put on its protection.
In this project, we propose a novel way to deal with save security of different partners engaged with the data brokering process. We are among the first to formally characterize two protection privacy attacks, namely attribute-correlation attack, and inference attack, and propose two countermeasure schemes automaton segmentation and query segment encryption to securely share the routing decision-making responsibility among a selected set of brokering servers. With comprehensive security analysis and experimental results, we show that our approach seamlessly integrates security enforcement with query routing to provide system-wide security with insignificant overhead
