AdvSGM: Differentially Private Graph Learning via Adversarial Skip-gram Model
This addresses privacy concerns for sensitive graph data in applications like social networks, though it is incremental as it builds on existing differential privacy and skip-gram methods.
The paper tackled the problem of privacy risks in graph embedding techniques by proposing AdvSGM, a differentially private skip-gram model for graphs using adversarial training, which achieved high data utility on six real-world datasets without additional noise injection.
The skip-gram model (SGM), which employs a neural network to generate node vectors, serves as the basis for numerous popular graph embedding techniques. However, since the training datasets contain sensitive linkage information, the parameters of a released SGM may encode private information and pose significant privacy risks. Differential privacy (DP) is a rigorous standard for protecting individual privacy in data analysis. Nevertheless, when applying differential privacy to skip-gram in graphs, it becomes highly challenging due to the complex link relationships, which potentially result in high sensitivity and necessitate substantial noise injection. To tackle this challenge, we present AdvSGM, a differentially private skip-gram for graphs via adversarial training. Our core idea is to leverage adversarial training to privatize skip-gram while improving its utility. Towards this end, we develop a novel adversarial training module by devising two optimizable noise terms that correspond to the parameters of a skip-gram. By fine-tuning the weights between modules within AdvSGM, we can achieve differentially private gradient updates without additional noise injection. Extensive experimental results on six real-world graph datasets show that AdvSGM preserves high data utility across different downstream tasks.