SGER: Trusted Privacy Preserving Data Mining over Grids

University Of Illinois At Chicago, Chicago IL

Investigators

Abstract

This project develops a service oriented middleware to facilitate the execution of secure fault tolerant multiparty computations in distributed collaborative applications over peer-to-peer networks. Salient tasks being pursued include: - Development of SMC application layer (algorithms/protocols) and associated services to facilitate seamless, efficient and scalable execution of SMC algorithms on large scale distributed overlay networks. - Design of efficient SMC Kernel functions integrated within the middleware which can serve the needs of a wide range of SMC applications, including aggregates, median filters, sorting, selection, graph computations, etc. - Efficient implementation of the algorithms and communication protocols in contemporary networked operating system environments. - Development of the layered middleware supporting trusted execution of SMC kernels. The project uses data mining and networking applications as a vehicle to motivate research issues and associated solutions. Apart from the importance of the research in advancing knowledge in the areas of secure multiparty computing, distributed algorithms, and grid computing, the project -- Facilitates pervasive deployment of grid environments, -- Motivates rapid developments of collaborations among commercial and government entities. -- Provides mechanism for cost effective access to high performance computing resources, and -- Helps in training graduate and undergraduate student on cutting-edge technology and prepare sthem to assume leadership role in this emerging field.