Privacy Preserving Data Mining and Data Exposure Technique and Performance Study

Abstract

The data mining is used in various applications in order to provide decisions, pattern matching and others. In order to mine data, it is required to initially learn from the data and then extract the patterns from data which is supplied in raw formats. Sometimes data is available in parts and need to conclude the outcomes from the data in secure manner. Therefore a technique is required that securely combine data from all the participating partitions of the data. Mine the combined data for extracting the meaningful pattern in a secured manner. Finally disclose the data in such manner which is utilized by the different other application, without disturbing the outcomes of the final application decisions. In order to design and demonstrate the three parties of data are considered for combining them, and mine them. And finally using the C4.5 classification algorithm is applied to find the final utilization on different application. To find that the mining decision is not varied from the initial values of mining the comparison is made between the disclosed data and initial dataset on the basis of accuracy of classification. The experimental results demonstrate a very fewer patterns are misclassified after in comparison of the initial dataset classification in terms of accuracy.