FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4378966525> ?p ?o ?g. }

• W4378966525 abstract "Graph Neural Networks (GNNs) have been widely applied to different tasks such as bioinformatics, drug design, and social networks. However, recent studies have shown that GNNs are vulnerable to adversarial attacks which aim to mislead the node (or subgraph) classification prediction by adding subtle perturbations. In particular, several attacks against GNNs have been proposed by adding/deleting a small amount of edges, which have caused serious security concerns. Detecting these attacks is challenging due to the small magnitude of perturbation and the discrete nature of graph data. In this paper, we propose a general adversarial edge detection pipeline EDoG without requiring knowledge of the attack strategies based on graph generation. Specifically, we propose a novel graph generation approach combined with link prediction to detect suspicious adversarial edges. To effectively train the graph generative model, we sample several sub-graphs from the given graph data. We show that since the number of adversarial edges is usually low in practice, with low probability the sampled sub-graphs will contain adversarial edges based on the union bound. In addition, considering the strong attacks which perturb a large number of edges, we propose a set of novel features to perform outlier detection as the preprocessing for our detection. Extensive experimental results on three real-world graph datasets including a private transaction rule dataset from a major company and two types of synthetic graphs with controlled properties (e.g., Erdos-Renyi and scale-free graphs) show that EDoG can achieve above 0.8 AUC against four state-of-the-art unseen attack strategies without requiring any knowledge about the attack type (e.g., degree of the target victim node); and around 0.85 with knowledge of the attack type. EDoG significantly outperforms traditional malicious edge detection baselines. We also show that an adaptive attack with full knowledge of our detection pipeline is difficult to bypass it. Our results shed light on several principles to improve the robustness of GNNs." @default.
• W4378966525 created "2023-06-02" @default.
• W4378966525 creator A5010398748 @default.
• W4378966525 creator A5028693742 @default.
• W4378966525 creator A5031954035 @default.
• W4378966525 creator A5039744812 @default.
• W4378966525 creator A5040046281 @default.
• W4378966525 date "2023-02-01" @default.
• W4378966525 modified "2023-09-23" @default.
• W4378966525 title "EDoG: Adversarial Edge Detection For Graph Neural Networks" @default.
• W4378966525 cites W1971937094 @default.
• W4378966525 cites W1983776999 @default.
• W4378966525 cites W2089554624 @default.
• W4378966525 cites W2101447063 @default.
• W4378966525 cites W2109394932 @default.
• W4378966525 cites W2124637492 @default.
• W4378966525 cites W2140584250 @default.
• W4378966525 cites W2162630660 @default.
• W4378966525 cites W2168190036 @default.
• W4378966525 cites W2420733993 @default.
• W4378966525 cites W2509798866 @default.
• W4378966525 cites W2514325387 @default.
• W4378966525 cites W2618043096 @default.
• W4378966525 cites W2772302578 @default.
• W4378966525 cites W2803831897 @default.
• W4378966525 cites W2809583854 @default.
• W4378966525 cites W2949208225 @default.
• W4378966525 cites W2963448658 @default.
• W4378966525 cites W2963969878 @default.
• W4378966525 cites W2964971928 @default.
• W4378966525 cites W2986933397 @default.
• W4378966525 cites W3104038788 @default.
• W4378966525 cites W3106390645 @default.
• W4378966525 cites W4289436753 @default.
• W4378966525 cites W4312512934 @default.
• W4378966525 doi "https://doi.org/10.1109/satml54575.2023.00027" @default.
• W4378966525 hasPublicationYear "2023" @default.
• W4378966525 type Work @default.
• W4378966525 citedByCount "0" @default.
• W4378966525 crossrefType "proceedings-article" @default.
• W4378966525 hasAuthorship W4378966525A5010398748 @default.
• W4378966525 hasAuthorship W4378966525A5028693742 @default.
• W4378966525 hasAuthorship W4378966525A5031954035 @default.
• W4378966525 hasAuthorship W4378966525A5039744812 @default.
• W4378966525 hasAuthorship W4378966525A5040046281 @default.
• W4378966525 hasBestOaLocation W43789665252 @default.
• W4378966525 hasConcept C119857082 @default.
• W4378966525 hasConcept C124101348 @default.
• W4378966525 hasConcept C132525143 @default.
• W4378966525 hasConcept C153180895 @default.
• W4378966525 hasConcept C154945302 @default.
• W4378966525 hasConcept C34736171 @default.
• W4378966525 hasConcept C37736160 @default.
• W4378966525 hasConcept C41008148 @default.
• W4378966525 hasConcept C80444323 @default.
• W4378966525 hasConceptScore W4378966525C119857082 @default.
• W4378966525 hasConceptScore W4378966525C124101348 @default.
• W4378966525 hasConceptScore W4378966525C132525143 @default.
• W4378966525 hasConceptScore W4378966525C153180895 @default.
• W4378966525 hasConceptScore W4378966525C154945302 @default.
• W4378966525 hasConceptScore W4378966525C34736171 @default.
• W4378966525 hasConceptScore W4378966525C37736160 @default.
• W4378966525 hasConceptScore W4378966525C41008148 @default.
• W4378966525 hasConceptScore W4378966525C80444323 @default.
• W4378966525 hasFunder F4320306076 @default.
• W4378966525 hasLocation W43789665251 @default.
• W4378966525 hasLocation W43789665252 @default.
• W4378966525 hasOpenAccess W4378966525 @default.
• W4378966525 hasPrimaryLocation W43789665251 @default.
• W4378966525 hasRelatedWork W1502614025 @default.
• W4378966525 hasRelatedWork W2066259560 @default.
• W4378966525 hasRelatedWork W2126100045 @default.
• W4378966525 hasRelatedWork W2262783296 @default.
• W4378966525 hasRelatedWork W2380927352 @default.
• W4378966525 hasRelatedWork W2391959412 @default.
• W4378966525 hasRelatedWork W2728578317 @default.
• W4378966525 hasRelatedWork W2801924399 @default.
• W4378966525 hasRelatedWork W3129710645 @default.
• W4378966525 hasRelatedWork W4211209597 @default.
• W4378966525 isParatext "false" @default.
• W4378966525 isRetracted "false" @default.
• W4378966525 workType "article" @default.