Privately learning statistics of events on devices can enable improved user experience. Differentially private algorithms for such problems can benefit significantly from interactivity. We argue that an aggregation protocol can enable an interactive private federated statistics system where user's devices maintain control of the privacy assurance. We describe the architecture of such a system, and analyze its security properties.
@article{arxiv.2211.10082,
title = {Private Federated Statistics in an Interactive Setting},
author = {Audra McMillan and Omid Javidbakht and Kunal Talwar and Elliot Briggs and Mike Chatzidakis and Junye Chen and John Duchi and Vitaly Feldman and Yusuf Goren and Michael Hesse and Vojta Jina and Anil Katti and Albert Liu and Cheney Lyford and Joey Meyer and Alex Palmer and David Park and Wonhee Park and Gianni Parsa and Paul Pelzl and Rehan Rishi and Congzheng Song and Shan Wang and Shundong Zhou},
journal= {arXiv preprint arXiv:2211.10082},
year = {2022}
}