Privacy-Preserving Mining of Association Rules in Cloud

With the appearance of cloud computing and its domains for IT services based on the internet and big data centers, the outsourcing of data and computing services is acquiring a huge relevance. The interest in the area of data mining, as a service, has been the main stay; because of the encouraged development in the various fields, such as Cloud Computing. A third party service provider, the server, comes in the frame, when a company, the data owner, who lacks in expertise or the resources, outsources its mining needs. However the data owner thinks both the items and the association rules of the outsourced Database as a confidential property. The server stores the data and ships transformed by the data owner. Then the data owner sends mining queries to the server, and the server returns the extracted patterns. From these patterns, the owner recovers the true patterns. Within corporate privacy-preserving frameworks, the problem of outsourcing the association rule mining responsibilities in the cloud environment is studied in this paper. An innovative approach ensures that every transformed item from data owner to server is interchangeable with compared to the background knowledge of attackers is proposed in the paper. Our methods are scalable, effectual and protect privacy on an extremely huge and real transaction database representative our complete algorithm. This approach also proposes to provide the privacy-preserving mining over cloud. We can assume that a conventional model where the adversary knows the area of items and their exact occurrence and can employ this information to identify cipher items and cipher item sets.