Robust meta network embedding against adversarial attacks
Document Type
Conference Proceeding
Publication Date
11-1-2020
Abstract
Recent studies have shown that graph mining models are vulnerable to adversarial attacks. This paper proposes a robust meta network embedding framework, RoMNE, which improves the robustness of multiple network embedding on adversarial noisy networks while preserving the utility on original clean ones. First, we propose a generic meta learning based multiple network embedding model that can quickly adapt it to new embedding tasks on a variety of network data with only a small number of parameter and training updates. Second, Gumbel estimator and Gaussian smoothing techniques are introduced to implement differentiable approximation for optimizing non-differential objective of effective adversarial attacks. Last but not least, the adversarial attack and defense models are integrated into a dynamic adversarial training model. The competition of two models helps the latter be robust to adversarial attacks.
Publication Source (Journal or Book title)
Proceedings - IEEE International Conference on Data Mining, ICDM
First Page
1448
Last Page
1453
Recommended Citation
Zhou, Y., Ren, J., Dou, D., Jin, R., Zheng, J., & Lee, K. (2020). Robust meta network embedding against adversarial attacks. Proceedings - IEEE International Conference on Data Mining, ICDM, 2020-November, 1448-1453. https://doi.org/10.1109/ICDM50108.2020.00192
