SECURE MEAN CALCULATION: AN EXAMPLE OF PRIVACY PRESERVING ANALYSIS

Data is increasing day by day. So we can't store them in local storage. But the cloud provides a huge resource, flexibility, and software. So we don't need to make a costly and more complicated system. On the other hand, because of handle various data mining issues, we need a statistical model of data. Statistical calculation tells us about the past and helps to predict our future. The most basic term in statistical information is the mean calculation. So to analyze the mean over large data, resource constraint clients take help from the cloud. But when we want to know about mean from large data stored in the cloud, there is always a security issue of data. Because the cloud is not fully trusted. So, for the security issue data is saved in cloud in encrypted form. So analysis can be invoked in encrypted data. But we want to know the actual analysis result. So the solution is the homomorphism encryption technique. For this, a central key generator is used to give an encrypted keys to clients and decrypted keys to the analyzer. Also, analyzer is not fully trusted. In this paper, we proposed a protocol to securely analyze client's data on the cloud by securely share private keys.