A cloud-based platform for encrypted data mining as a service

The accelerated production of data in digital format has intensified the need for efficient data analysis and knowledge extraction tools for decision-making. With the increasing use of cloud storage due to the vast amount of data generated every day, the outsourcing of data analytic to third parties is also becoming a necessity. However, when the nature of the data is sensitive, security and privacy issues need to be addressed as part of the outsourcing process as the service provider can act honestly but curiously. This paper presents a cloud-based platform that enable privacy-preserving data mining tasks as a service. The platform provides the tools for the data owner and the untrusted service provider to efficiently execute privacy preserving data mining (PPDM) schemes in the context of Big data. As a case study, the platform was validated and evaluated using PPDMs relying on homomorphic encryption for secure clustering. From the experimental evaluation, the results revealed the feasibility of data mining as a service with data privacy preserving and without accuracy loss.