SECURING HEALTH CARE DATA IN COLLABORATIVE DATA PUBLISHING USING MAPREDUCE FRAMEWORK
Shital Suryawanshi, Vinod S. Wadne · 2015
Today publishing data on a web becomes need. Publishing such microdata can breach privacy of any individual. For analysis purpose researcher, medical practitioner required such health related data. Existing system used encryption algorithms for securing data. The data is stored on HDFS by encrypting and the user having key can only access that data by decrypting it. Big data is heterogeneous, distributed data where data is collected from different sources having different dimensions. In hospital patients data can be stored in different form such as audio, video, and in images. Big data having different characteristics like variety of data, its volume and velocity, this makes it different from other databases. Data privacy is one of the challenge in data mining with big data. Big data keeps growing continuously. In case of big data it is not efficient to encrypt large amount of data as it is time consuming. Existing provider aware algorithm has problem of data loss due to insider attack. K-anonymity and l- diversity are very popular algorithms for generalization and bucketization. They have some their own little limitations. In insider attack provider can infer the information of other user using his own records and with some background knowledge. To preserving the privacy of the user we need to use some method so that data privacy is preserve and at the same time increase the data utility. In the proposed system we focus to maintain the privacy for distributed data, and overcome the problems of M-privacy using new updated provider algorithm with a slicing technique. The main goal of paper is to publish an anonymized view of integrated data, which will be immune to attacks. We also use MR-Cube method which is used to compute large cube with non algebraic measures such as TOP-k, count.