Privacy preserving data sharing and analysis for edge-based architectures

In this paper, we present a framework for privacy preserving collaborative data analysis among multiple data providers acting as edge of a cloud environment. The proposed framework computes the best trade-off among privacy and result accuracy, based on the privacy requirements of data providers and the specific requested analysis algorithm. Though the presented model is general and can be applied to different environments, this work is motivated by the need of sharing information related to Cyber Threats (CTI). The presented framework is independent from the number of data providers, used data format, privacy requirement and analysis operations. The model is based on the concepts of trade-off score between accuracy and privacy, which also considers measures for privacy requirement such as differential privacy, l-diversity and k-anonymity. Together with the model, the paper discusses the framework implementation and presents results to show the effectiveness and viability of the proposed approach.