Available Protocols for Multi-party Computation in Privacy Preserving Data Mining

Abstract

Privacy preservation is an important component in designing various data mining applications to seek a trade-off between mining utilities and protecting private information. Due to vast advancements in internet, computing and communication, multiparty computation has great opportunities to perform cooperative computation on multiple parties. In multiparty computation, set of parties with similar background wants to collaboratively compute some function on their private inputs. This major challenge on researchers is how to protect privacy of their sensitive input data without disclosing it to others. Privacy preservation is done by Secure Multiparty Computation, Data Perturbation, Anonymization and Role based access control in multiparty environments. Several protocols have been developed for privacy preserving multiparty computation. These protocols are used to perform multiparty collaborative data mining to preserve privacy of data and knowledge when common users are involved in data mining. This paper describes various protocols for Secure Multiparty Computation and Data Perturbation along with certain security properties like correctness and privacy. Finally the comparisons of various protocols are also discussed.