An Efficient Approach for Privacy Preserving Distributed K-Means Clustering Based on Shamir’s Secret Sharing Scheme

摘要: Privacy preserving data mining has gained considerable attention because of the increased concerns to ensure privacy sensitive information. Amongst two basic approaches for mining, viz. Randomization based and Cryptography based, later provides high level but incurs higher computational as well communication overhead. Hence, it is necessary explore alternative techniques that improve overheads. In this work, we propose an efficient, collusion-resistant cryptography approach distributed K-Means clustering using Shamir’s secret sharing scheme. As show from theoretical practical analysis, our provably secure does not require a trusted third party. addition, negligible overhead compared existing approaches.