English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Graph neural networks (GNNs) have achieved state-of-the-art performance inmany graph-based tasks such as node classification and graph classification.However, many recent works have demonstrated that an attacker can mislead GNNmodels by slightly perturbing the graph structure. Existing attacks to GNNs areeither under the less practical threat model where the attacker is assumed toaccess the GNN model parameters, or under the practical black-box threat modelbut consider perturbing node features that are shown to be not enougheffective. In this paper, we aim to bridge this gap and consider black-boxattacks to GNNs with structure perturbation as well as with theoreticalguarantees. We propose to address this challenge through bandit techniques.Specifically, we formulate our attack as an online optimization with banditfeedback. This original problem is essentially NP-hard due to the fact thatperturbing the graph structure is a binary optimization problem. We thenpropose an online attack based on bandit optimization which is proven to be{sublinear} to the query number $T$, i.e., $\mathcal{O}(\sqrt{N}T^{3/4})$ where$N$ is the number of nodes in the graph. Finally, we evaluate our proposedattack by conducting experiments over multiple datasets and GNN models. Theexperimental results on various citation graphs and image graphs show that ourattack is both effective and efficient. Source code is availableat~\url{https://github.com/Metaoblivion/Bandit_GNN_Attack}

Bandits for Structure Perturbation-based Black-box Attacks to Graph Neural Networks with Theoretical Guarantees