Securing Big Data Efficiently through Microaggregation Technique

Cyber-Physical Systems (CPS) requires big data communications as well as integration from several distributed sources. This data can usually be interconnected with physical applications, such as power grids or SCADA systems. In addition, it can be publicly accessible for using by third party users or data scientists. Therefore, it becomes imperative to ensure that this big data is well secured. Microaggregation is an widely used technique to protect a dataset through anonymity in order to prevent exposure of a person's identity. This data disclosure may also result from an unpredicted data linkage with another dataset. As, most of these survey datasets store records using numerical values, many of the microaggregation techniques are developed and tested on numerical data. These algorithms are not suitable for those data where both numerical and categorical data are stored. In this paper we're proposing a microaggregation technique in order to provide data anonymity regardless of its type. The records are clustered into several groups using an evolutionary attribute grouping algorithm and each group records are then microaggregated applying Huffman data compression algorithm.