Data Poisoning Attack against Unsupervised Node Embedding Methods
By Mingjie Sun, Jian Tang, Huichen Li, Bo Li, Chaowei Xiao, Yao Chen, Dawn Song. 2018
70 citations recorded by the source corpus.
Unsupervised node embedding methods (e.g., DeepWalk, LINE, and node2vec) have attracted growing interests given their simplicity and effectiveness. However, although these methods have been proved effective in a variety of applications, none of the existing work has analyzed the robustness of them. This could be very risky if these methods are attacked by an adversarial party. In this paper, we take the task of link prediction as an example, which is one of the most fundamental problems for graph analysis, and introduce a data positioning attack to node embedding methods. We give a complete characterization of attacker's utilities and present efficient solutions to adversarial attacks for two popular node embedding methods: DeepWalk and LINE. We evaluate our proposed attack model on multiple real-world graphs. Experimental results show that our proposed model can significantly affect the results of link prediction by slightly changing the graph structures (e.g., adding or removing a few edges). We also show that our proposed model is very general and can be transferable across different embedding methods. Finally, we conduct a case study on a coauthor network to better understand our attack method.
Subfields: Artificial Intelligence, Computer Networks and Communications
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Neil Zhenqiang Gong
- Binghui Wang
- Alan L. Yuille
- Philip S. Yu
- Lichao Sun
- Xiaohui Shen
- Bo Li
- Cihang Xie
In Computer Networks and Communications
- Binghui Wang
- Neil Zhenqiang Gong
- Bo Li
- Philip S. Yu
- Alan L. Yuille
- Evangelos E. Papalexakis
- Lichao Sun
- Yingtong Dou
Citations in the corpus
- Attacking Graph-based Classification via Manipulating the Graph Structure (2019, methodology): “Several recent work [16,55,74] proposed attacks to graph neural network methods via manipulating the graph structure, i.e., inserting fake edges to the graph or deleting existing edges.”
- Adversarial Attack and Defense on Graph Data: A Survey (2018, background): “Citation Network Node/Link Citeseer [87] 3,327 4,732 3,703 6 [145], [28], [23], [106], [92], [8], [147], [12], [129], [120], [125], [17], [127], [40], [140], [89], [22], [144], [105] , [57], [148], [9], [52], [53], [107], [36], [134], [96], [20], [54], [29], [56], [31], [76], [77], [146], [59], [49]”
- Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses (2019, background): “The variants of adversarial attacks [3, 9, 12, 20, 21, 30, 32, 33, 53, 56, 63, 64, 80] cast a security threat when deploying machine learning systems.”
- All You Need Is Low (Rank): Defending Against Adversarial Attacks on Graphs (2020, background): “The other group of studies, investigate the effects of adversarial attacks on unsupervised node embeddings [7, 35, 38].”
- Humpty Dumpty: Controlling Word Meanings via Corpus Poisoning (2020, unknown): “Gradient-based poisoning attacks on matrix factorization have been suggested in the context of collaborative filtering [43] and adapted to unsupervised node embeddings [68].”