Secure Bayesian Federated Analytics for Privacy-Preserving Trend Detection
This work addresses privacy concerns in federated analytics for edge computing applications, offering a method that could enhance decision-making in services and products, though it appears incremental as it builds on existing Bayesian and secure aggregation techniques.
The paper tackles the problem of privacy-preserving trend detection in federated analytics by proposing a Bayesian approach with a secure aggregation protocol called SAFE, which reduces computational burden for users and aggregators while maintaining privacy for production use cases.
Federated analytics has many applications in edge computing, its use can lead to better decision making for service provision, product development, and user experience. We propose a Bayesian approach to trend detection in which the probability of a keyword being trendy, given a dataset, is computed via Bayes' Theorem; the probability of a dataset, given that a keyword is trendy, is computed through secure aggregation of such conditional probabilities over local datasets of users. We propose a protocol, named SAFE, for Bayesian federated analytics that offers sufficient privacy for production grade use cases and reduces the computational burden of users and an aggregator. We illustrate this approach with a trend detection experiment and discuss how this approach could be extended further to make it production-ready.