Achieving K-Anonymity Using Parallelism in Full Domain Generalization

Preserving privacy while publishing data has emerged as key research area in data security and has become a primary issue in publishing person specific sensitive information. How to preserve one’s privacy efficiently is a critical issue while publishing data. k - Anonymity is a key technique for de-identifying the sensitive datasets. In our work, we have described an approach to implement various k anonymity algorithms and also propose a parallelism method that produces better results with the real-world datasets. Additionally, we suggest a new approach that attains better results by applying a parallelism approach and exploiting various characteristics of our suggested approach. The proposed approach uses the concept of samarati algorithm to generalize the lattice and uses the binary search method. The proposed algorithm generates the levels using binary search in the lattice and then uses the parallel mechanism for evaluating the nodes. The proposed algorithm has less execution time than other full domain generalization algorithms for k -anonymization