A Privacy Preserving Scheme for Big data Publishing in the Cloud using k-Anonymization and Hybridized Optimization Algorithm

摘要

One of the emerging research areas in the recent years is big data due to the enormous data flow in various fields, like hospitals, government records, social sites, etc. In this field, cloud computing has drawn significant importance as the user can transfer huge volume of data through the servers. Hence, it is necessary to protect the data so that the third party cannot access the information provided by the cloud users. This work introduces the k-anonymization model for privacy preservation in the cloud. The proposed scheme is driven by the newly developed optimization model, namely Dragon Particle Swarm Optimization (Dragon-PSO) which combines the Dragonfly Algorithm (DA) and Particle Swarm Optimization (PSO) algorithm. The proposed scheme derives the fitness function for the proposed Dragon-PSO algorithm attaining high value for privacy and utility. The proposed scheme is evaluated based on two metrics, Information Loss and Classification Accuracy.