Privacy Based Association Rules in Secure Horizontal Database

Privacy in data mining prevents the parties from directly sharing the data, and some types of information about the data. We introduce a protocol for secure mining of association rules in horizontally distributed databases using FDM, which an unsecured distributed version of the Apriori algorithm. We extended our work with Horizontal Aggregation that give rise to multiple row output, which transform rows to column using CASE, SPJ or PIVOT operators depending on the input. In order to prepare real world datasets that are very much suitable for data mining operations, we explored horizontal aggregations by developing constructs in the form of operators such as CASE, SPJ and PIVOT. Instead of single value, the horizontal aggregations return a set of values in the form of a row. The proposed system is user friendly as users are never expected to write queries we introduced security and privacy through cryptography and level based slicing. Level Based Slicing introduces different techniques such as Generalized Data, Bucketized Data, Multiset-based Generalization Data, One-attribute-per-Column Slicing Data, Sliced Data dimensionality of the data and preserves better utility than generalization and bucketization. Slicing protects privacy because it breaks the associations between uncorrelated attributes, which are infrequent and thus identifying. Our result shows that proposed method have higher performance other sequential algorithms. Customers can dynamically send request to the server either to apply privacy and security. In privacy, server can apply five types of slicing methods. In security server can apply FDM, Encryption and decryption, Hash function generate key for encryption and decryption. Experimental Analysis can be comparing with the time taken to evaluate the privacy preserving methods and Enhanced FDM Algorithm