Russian company Big Data Platform (BDP), a joint venture by national long-distance telephone provider Rostelecom and the country's bank VTB, has completed tests of the new Big Data security platform.
Multi-Partial Computations (MPC) technology enables companies to access shared use of Big Data and raise their business efficiency in this way.
"The characteristic of MPC technology is the absence of the exchange of source data between companies, which allows service users to implement the Data Fusion approach. This consists of combining data and merging or transferring algorithms from one area of machine learning (ML) to another, as well as the integration of machine learning processes," said a press release from Big Data Platform.
The Multi-Partial Computations technology was tested with the participation of Ubic, a traffic filtering developer and analysis service provider. Collaborative confidential computing allows users to create machine learning models that are built on the combined data of various companies.
Each party shares information and transmits only the amount of protected data on which models can be built. At the same time, it is impossible to decipher the information for obtaining primary data. This technology allows users to develop services based on various sources and helps enrich the competencies and expertise of various areas, from retail to the insurance business.
Maxim Konovalikhin, Senior Vice President, Head of the Data Analysis and Modeling Department at VTB, said: "The security of customer data is the priority of the Big Data Platform; the new technology allows us to bring our work with Big Data and data security to a new level. Thanks to MPC, we can work with combined arrays of information to improve products, services, and business processes. Combining unique arrays of depersonalized data, technologies, and competencies, as well as applying the Data Fusion approach, will allow the Big Data Platform to achieve leadership in the Big Data market.”