公开(公告)号：US20250068658A1

公开(公告)日：2025-02-27

申请号：US18941263

申请日：2024-11-08

Applicant: SAS Institute Inc.

Inventor： Kai Xu ,  Georgi Valentinov Ganev ,  Emile Isak Joubert ,  Rees Stephen Davison ,  Olivier Rene Maurice Van Acker ,  Luke Anthony William Robinson ,  Sofiane Mahiou

IPC: G06F16/28

Abstract: Embodiments described herein relate to the efficient generation of synthetic datasets that represent many-to-many relationships. In particular, certain embodiments implement a particular factorization for many-to-many generative models, which leads to a scalable generation framework by combining random graph theory and representation learning. Further embodiments we extend the framework to establish the notion of differential privacy within the synthetically generated data. The embodiments described herein are therefore able to generate synthetic datasets efficiently while preserving information within and across many-to-many datasets with improved accuracy.