Graph Datasets Research Articles

Design your own universe: a physics-informed agnostic method for enhancing graph neural networks

AbstractPhysics-informed Graph Neural Networks have achieved remarkable performance in learning through graph-structured data by mitigating common GNN challenges such as over-smoothing, over-squashing, and heterophily adaption. Despite these advancements, the development of a simple yet effective paradigm that appropriately integrates previous methods for handling all these challenges is still underway. In this paper, we draw an analogy between the propagation of GNNs and particle systems in physics, proposing a model-agnostic enhancement framework. This framework enriches the graph structure by introducing additional nodes and rewiring connections with both positive and negative weights, guided by node labeling information. We theoretically verify that GNNs enhanced through our approach can effectively circumvent the over-smoothing issue and exhibit robustness against over-squashing. Moreover, we conduct a spectral analysis on the rewired graph to demonstrate that the corresponding GNNs can fit both homophilic and heterophilic graphs. Empirical validations on benchmarks for homophilic, heterophilic graphs, and long-term graph datasets show that GNNs enhanced by our method significantly outperform their original counterparts.

NLA-GNN: Non-local information aggregated graph neural network for heterogeneous graph embedding

Heterogeneous graphs are ubiquitous in the real world. Recent methods aim to obtain meaningful low-dimensional node embeddings from heterogeneous graphs to facilitate downstream applications. However, most existing methods tend to consider the local information but ignore the non-local information. This paper proposes a novel Non-Local Information Aggregated Graph Neural Network (NLA-GNN) that aggregate not only the local information from neighbor nodes but also non-local information from distant nodes. Specifically, Local aggregation modules in NLA-GNN utilize the attention mechanism to generate potentially valuable metapaths and use them to aggregate local information. In contrast, non-local aggregation modules adopt a two-step approach, and each step uses an attention-guided method to sort nodes into node sequences and aggregate information with the methods designed for sequential data. Experiment results on three heterogeneous graph datasets demonstrate the performance of NLA-GNN over state-of-the-arts and the necessity of non-local aggregation in heterogeneous graphs.

Memristive Crossbar Array-Based Probabilistic Graph Modeling.

Modern graph datasets with structural complexity and uncertainties due to incomplete information or data variability require advanced modeling techniques beyond conventional graph models. This study introduces a memristive crossbar array (CBA)-based probabilistic graph model (C-PGM) utilizing Cu0.3Te0.7/HfO2/Pt memristors, which exhibit probabilistic switching, self-rectifying, and memory characteristics. C-PGM addresses the complexities and uncertainties inherent in structural graph data across various domains, leveraging the probabilistic nature of memristors. C-PGM relies on the device-to-device variation across multiple memristive CBAs, overcoming the limitations of previous approaches that rely on sequential operations, which are slower and have a reliability concern due to repeated switching. This new approach enables the fast processing and massive implementation of probabilistic units at the expense of chip area. In this study, the hardware-based C-PGM feasibly expresses small-scale probabilistic graphs and shows minimal error in aggregate probability calculations. The probability calculation capabilities of C-PGM are applied to steady-state estimation and the PageRank algorithm, which is implemented on a simulated large-scale C-PGM. The C-PGM-based steady-state estimation and PageRank algorithm demonstrate comparable accuracy to conventional methods while significantly reducing computational costs.

Affinity Uncertainty-Based Hard Negative Mining in Graph Contrastive Learning.

Hard negative mining has shown effective in enhancing self-supervised contrastive learning (CL) on diverse data types, including graph CL (GCL). The existing hardness-aware CL methods typically treat negative instances that are most similar to the anchor instance as hard negatives, which helps improve the CL performance, especially on image data. However, this approach often fails to identify the hard negatives but leads to many false negatives on graph data. This is mainly due to that the learned graph representations are not sufficiently discriminative due to oversmooth representations and/or non-independent and identically distributed (non-i.i.d.) issues in graph data. To tackle this problem, this article proposes a novel approach that builds a discriminative model on collective affinity information (i.e., two sets of pairwise affinities between the negative instances and the anchor instance) to mine hard negatives in GCL. In particular, the proposed approach evaluates how confident/uncertain the discriminative model is about the affinity of each negative instance to an anchor instance to determine its hardness weight relative to the anchor instance. This uncertainty information is then incorporated into the existing GCL loss functions via a weighting term to enhance their performance. The enhanced GCL is theoretically grounded that the resulting GCL loss is equivalent to a triplet loss with an adaptive margin being exponentially proportional to the learned uncertainty of each negative instance. Extensive experiments on ten graph datasets show that our approach does the following: 1) consistently enhances different state-of-the-art (SOTA) GCL methods in both graph and node classification tasks and 2) significantly improves their robustness against adversarial attacks. Code is available at https://github.com/mala-lab/AUGCL.

MDD-FedGNN: A vertical federated graph learning framework for malicious domain detection

The domain name system (DNS) serves as a fundamental component of the Internet infrastructure, but it is also exploited by attackers in various cyber-crimes, underscoring the significance of malicious domain detection (MDD). Recent advances show that graph-based models exhibit potential for inferring malicious domains and demonstrate superior performance. However, acquiring large-scale and high-quality graph datasets for MDD proves challenging for individual security institutes. Hence, a promising research direction involves employing vertical federated graph learning scheme to unite diverse security institutes and enhance local datasets resulting in more robust and powerful detection models. Nonetheless, directly applying vertical federated graph neural networks for MDD confronts challenges posed by noisy labels and noisy edges among security institutes, which ultimately diminish detection performance. This paper introduces a novel vertical federated learning framework, called MDD-FedGNN, that applies contrastive learning with two different encoders to deal with noisy labels and employs a new loss function based on the information bottleneck theory to handle noisy edges. Comparative experiments are conducted on a publicly available DNS dataset to evaluate the effectiveness of MDD-FedGNN in addressing the challenges of noisy labels and edges in vertical federated graph learning. The results demonstrate that MDD-FedGNN outperforms baseline methods, confirming the feasibility of training more powerful malicious domain detection models through data sharing and vertical federated learning among different security agencies.

Hybrid Quantum or Purely Classical? Assessing the Utility of Quantum Feature Embeddings

Background As graph datasets—including social networks, supply chains, and bioinformatics data—grow in size and complexity, researchers are driven to search for solutions enhancing model efficiency and speed. One avenue that may provide a solution is Quantum Graph Learning (QGL), a subfield of Quantum Machine Learning (QML) that applies machine learning inspired or powered by quantum computing to graph learning tasks. Methods We reevaluate Quantum Feature Embeddings (QFE), a QGL methodology published by Xu et al. earlier this year. QFE uses Variational Quantum Circuits to preprocess node features and then sends them to a classical Graph Neural Network (GNN), with the goal of increasing performance and/or decreasing total model size. Xu et al. evaluated this methodology by comparing its performance with the performance of variously-sized classical models on the benchmark datasets PROTEINS and ENZYMES, and they report success. Our core methodology and learning task remain unchanged. However, we have made several changes to the experimental design that enhance the rigor of the study: 1) we include the testing of models with no embedder; 2) we conduct a thorough hyperparameter search using a state-of-the-art optimization algorithm; and 3) we conduct stratified five-fold cross-validation, which mitigates the bias produced by our small datasets and provides multiple test statistics from which we can calculate a confidence interval. Results We produce classical models that perform comparably to QFE and significantly outperform the small classical models used in Xu et al.’s comparison. Notably, many of our classical models achieve this using fewer parameters than the QFE models we trained. Xu et al. do not report their total model sizes. Conclusion Our study sheds doubt on the efficacy of QFE by demonstrating that small, well-tuned classical models can perform just as well as QFE, highlighting the importance of hyperparameter tuning and rigorous experimental design.

Graph classification using high-difference-frequency subgraph embedding

With the rapid growth of big data analysis, graphs have become an important data structure in relationship extraction and learning. However, the complexity of graph structure increases the difficulty of downstream learning tasks, e.g., graph classification. For example, some traditional graph classification methods rely on the attributes of nodes and edges; the attributes may be incomplete or missing in large-scale graph datasets, e.g., online social networks and knowledge graphs. Focusing on the node-level features overlooks the global characteristics, leading to decreased classification accuracy. This paper proposes a novel graph classification method based on the subgraph-level feature the high-difference-frequency subgraph. We measure the difference in subgraph appearing frequency and use the high difference frequency subgraphs as features in graph learning and classification. The experiments demonstrate that the proposed method outperforms the state-of-the-art graph classification methods.

Data Pruning-enabled High Performance and Reliable Graph Neural Network Training on ReRAM-based Processing-in-Memory Accelerators

Graph Neural Networks (GNNs) have achieved remarkable accuracy in cognitive tasks such as predictive analytics on graph-structured data. Hence, they have become very popular in diverse real-world applications. However, GNN training with large real-world graph datasets in edge-computing scenarios is both memory- and compute-intensive. Traditional computing platforms such as CPUs and GPUs do not provide the energy efficiency and low latency required in edge intelligence applications due to their limited memory bandwidth. Resistive random-access memory (ReRAM)-based processing-in-memory (PIM) architectures have been proposed as suitable candidates for accelerating AI applications at the edge, including GNN training. However, ReRAM-based PIM architectures suffer from low reliability due to their limited endurance, and low performance when they are used for GNN training in real-world scenarios with large graphs. In this work, we propose a learning-for-data-pruning framework, which leverages a trained Binary Graph Classifier (BGC) to reduce the size of the input data graph by pruning subgraphs early in the training process to accelerate the GNN training process on ReRAM-based architectures. The proposed light-weight BGC model reduces the amount of redundant information in input graph(s) to speed up the overall training process, improves the reliability of the ReRAM-based PIM accelerator, and reduces the overall training cost. This enables fast, energy-efficient, and reliable GNN training on ReRAM-based architectures. Our experimental results demonstrate that using this learning for data pruning framework, we can accelerate GNN training and improve the reliability of ReRAM-based PIM architectures by up to 1.6×, and reduce the overall training cost by 100× compared to state-of-the-art data pruning techniques.

TransE-MTP: A New Representation Learning Method for Knowledge Graph Embedding with Multi-Translation Principles and TransE

The purpose of representation learning is to encode the entities and relations in a knowledge graph as low-dimensional and real-valued vectors through machine learning technology. Traditional representation learning methods like TransE, a method which models relationships by interpreting them as translations operating on the low-dimensional embeddings of a graph’s entities, are effective for learning the embeddings of knowledge bases, but struggle to effectively model complex relations like one-to-many, many-to-one, and many-to-many. To overcome the above issues, we introduce a new method for knowledge representation, reasoning, and completion based on multi-translation principles and TransE (TransE-MTP). By defining multiple translation principles (MTPs) for different relation types, such as one-to-one and complex relations like one-to-many, many-to-one, and many-to-many, and combining MTPs with a typical translating-based model for modeling multi-relational data (TransE), the proposed method, TransE-MTP, ensures that multiple optimization objectives can be targeted and optimized during training on complex relations, thereby providing superior prediction performance. We implement a prototype of TransE-MTP to demonstrate its effectiveness at link prediction and triplet classification on two prominent knowledge graph datasets: Freebase and Wordnet. Our experimental results show that the proposed method enhanced the performance of both TransE and knowledge graph embedding by translating on hyperplanes (TransH), which confirms its effectiveness and competitiveness.

Graph Relearn Network: Reducing performance variance and improving prediction accuracy of graph neural networks

Recent studies show that the predictive performance of graph neural networks (GNNs) is inconsistent and varies across different experimental runs, even with identical parameters. The prediction variability limits GNNs’ applicability, and the underlying reasons remain unclear. We identified a key factor contributing to this issue: the oscillation of the predicted classes of some nodes during GNN training. To address this problem, we propose a novel framework, known as Graph Relearn Network (GRN), designed to reduce prediction variance by iteratively refining the predictions of unstable nodes. The GRN framework operates in two phases: pre-predict and relearn. During the pre-predict phase, a graph-dense encoder is trained to pre-predict the node categories. In the relearn phase, the model intensively focuses on the unstable nodes to optimize the predictions. Extensive experiments on ten graph datasets demonstrate that the GRN significantly enhances the performance stability of GNNs (with std. reduced by up to 75%), and achieves state-of-the-art performance in prediction accuracy (increased by up to 11.97%). By mitigating the instability introduced by unstable nodes, GRN enhances both the performance stability and prediction accuracy of GNNs in node classification tasks.

Sparse graphs-based dynamic attention networks

In previous research, the prevailing assumption was that Graph Neural Networks (GNNs) precisely depicted the interconnections among nodes within the graph's architecture. Nonetheless, real-world graph datasets are often rife with noise, elements that can disseminate through the network and ultimately affect the outcome of the downstream tasks. Facing the complex fabric of real-world graphs and the myriad potential disturbances, we introduce the Sparse Graph Dynamic Attention Networks (SDGAT) in this research. SDGAT employs the L0 regularization technique to achieve a sparse representation of the graph structure, which eliminates noise and generates a more concise sparse graph. Building upon this foundation, the model integrates a dynamic attention mechanism, allowing it to selectively focus on key nodes and edges, filter out irrelevant data, and simultaneously facilitate effective feature aggregation with important neighbors. To evaluate the performance of SDGAT, we conducted experiments on three citation datasets and compared its performance against commonly employed models. The outcomes indicate that SDGAT excels in node classification tasks, notably on the Cora dataset, with an accuracy rate of 85.29%, marking a roughly 3% enhancement over the majority of baseline models. The experimental findings provide evidence that SDGAT delivers effective performance on all three citation datasets, underscoring the efficacy of the dynamic attention network built upon a sparse graph.

Fair-RGNN: Mitigating Relational Bias on Knowledge Graphs

Knowledge graph data are prevalent in real-world applications, and knowledge graph neural networks (KGNNs) are essential techniques for knowledge graph representation learning. Although KGNN effectively models the structural information from knowledge graphs, these frameworks amplify the underlying data bias that leads to discrimination towards certain groups or individuals in resulting applications. Additionally, as existing debiasing approaches mainly focus on entity-wise bias, eliminating the multi-hop relational bias that pervasively exists in knowledge graphs remains an open question. However, it is very challenging to eliminate relational bias due to the sparsity of the paths that generate the bias and the non-linear proximity structure of knowledge graphs. To tackle the challenges, we propose Fair-KGNN, a KGNN framework that simultaneously alleviates multi-hop bias and preserves the proximity information of entity-to-relation in knowledge graphs. The proposed framework is generalizable to mitigate relational bias for all types of KGNN. Fair-KGNN is applicable to incorporate two stateof- the-art KGNN models, RGCN and CompGCN, to mitigate gender-occupation and nationality-salary bias. The experiments carried out on three benchmark knowledge graph datasets demonstrate that Fair-KGNN can effectively mitigate unfair situations during representation learning while preserving the predictive performance of KGNN models. The source code of the proposed method is available at: https://github.com/ynchuang/Mitigating-Relational-Bias-on-Knowledge-Graphs .

ADPSCAN: Structural Graph Clustering with Adaptive Density Peak Selection and Noise Re-Clustering

Structural graph clustering is a data analysis technique that groups nodes within a graph based on their connectivity and structural similarity. The Structural graph clustering SCAN algorithm, a density-based clustering method, effectively identifies core points and their neighbors within areas of high density to form well-defined clusters. However, the clustering quality of SCAN heavily depends on the input parameters, ϵ and μ, making the clustering results highly sensitive to parameter selection. Different parameter settings can lead to significant differences in clustering results, potentially compromising the accuracy of the clusters. To address this issue, a novel structural graph clustering algorithm based on the adaptive selection of density peaks is proposed in this paper. Unlike traditional methods, our algorithm does not rely on external parameters and eliminates the need for manual selection of density peaks or cluster centers by users. Density peaks are adaptively identified using the generalized extreme value distribution, with consideration of the structural similarities and interdependencies among nodes, and clusters are expanded by incorporating neighboring nodes, enhancing the robustness of the clustering process. Additionally, a distance-based structural similarity method is proposed to re-cluster noise nodes to the correct clusters. Extensive experiments on real and synthetic graph datasets validate the effectiveness of our algorithm. The experiment results show that the ADPSCAN has a superior performance compared with several state-of-the-art (SOTA) graph clustering methods.

Towards Exploring the Limitations of Test Selection Techniques on Graph Neural Networks: An Empirical Study

Graph Neural Networks (GNNs) have gained prominence in various domains, such as social network analysis, recommendation systems, and drug discovery, due to their ability to model complex relationships in graph-structured data. GNNs can exhibit incorrect behavior, resulting in severe consequences. Therefore, testing is necessary and pivotal. However, labeling all test inputs for GNNs can be prohibitively costly and time-consuming, especially when dealing with large and complex graphs. In response to these challenges, test selection has emerged as a strategic approach to alleviate labeling expenses. The objective of test selection is to select a subset of tests from the complete test set. While various test selection techniques have been proposed for traditional deep neural networks (DNNs), their adaptation to GNNs presents unique challenges due to the distinctions between DNN and GNN test data. Specifically, DNN test inputs are independent of each other, whereas GNN test inputs (nodes) exhibit intricate interdependencies. Therefore, it remains unclear whether DNN test selection approaches can perform effectively on GNNs. To fill the gap, we conduct an empirical study that systematically evaluates the effectiveness of various test selection methods in the context of GNNs, focusing on three critical aspects: 1) Misclassification detection: selecting test inputs that are more likely to be misclassified; 2) Accuracy estimation: selecting a small set of tests to precisely estimate the accuracy of the whole testing set; 3) Performance enhancement: selecting retraining inputs to improve the GNN accuracy. Our empirical study encompasses 7 graph datasets and 8 GNN models, evaluating 22 test selection approaches. Our study includes not only node classification datasets but also graph classification datasets. Our findings reveal that: 1) In GNN misclassification detection, confidence-based test selection methods, which perform well in DNNs, do not demonstrate the same level of effectiveness; 2) In terms of GNN accuracy estimation, clustering-based methods, while consistently performing better than random selection, provide only slight improvements; 3) Regarding selecting inputs for GNN performance improvement, test selection methods, such as confidence-based and clustering-based test selection methods, demonstrate only slight effectiveness; 4) Concerning performance enhancement, node importance-based test selection methods are not suitable, and in many cases, they even perform worse than random selection.

Diverse joint nonnegative matrix tri-factorization for attributed graph clustering

Cluster analysis of attributed graphs is a demanding and challenging task in the analysis of network-structured data. It involves learning node representation by leveraging both node attributes and the topological structure of the graph, aiming to accomplish effective clustering. Typically, existing methods fuse the topological and non-topological information by learning a consensus representation, often resulting in redundancy and overlooking their inherent distinctions. To address this issue, this paper proposes the Diverse Joint Nonnegative Matrix Tri-Factorization (Div-JNMTF), an embedding based model to detect communities in attributed graphs. The novel JNMTF model attempts to extract two distinct node representations from topological and non-topological data. Simultaneously, a diversity regularization technique utilizing the Hilbert–Schmidt Independence Criterion (HSIC) is employed. Its objective is to reduce redundant information in the node representations while encouraging the distinct contributions of both types of information. In addition, two graph regularization terms are introduced to preserve the local structures in the topological and attribute representation spaces. The Div-JNMTF model is optimized by developing an iterative optimization approach. By conducting thorough experiments on four synthetic and eight real-world attributed graph datasets, it has been demonstrated that the proposed model excels in accurately detecting attributed communities and surpasses the performance of existing methods.

Edge Deletion based Subgraph Hiding

Extracting subgraphs from graph data is a challenging and important subgraph mining task since they reveal valuable insights in many domains. However, in the data sharing scenario, some of the subgraphs might be considered as sensitive by the data owner and require hiding before publishing the data. Therefore, subgraph hiding is applied to the data so that when subgraph mining algorithms, such as frequent subgraph mining, subgraph counting, or subgraph matching, are executed on this published data, sensitive subgraphs will not appear. While protecting the privacy of the sensitive subgraphs through hiding, the side effects should be kept at a minimum. In this paper, we address the problem of hiding sensitive subgraphs on graph data and propose an Edge deletion-based heuristic (EDH) algorithm. We evaluate our algorithm using three graph datasets and compare the results with the previous vertex masking heuristic algorithms in terms of execution time and side effects in the context of frequent subgraph hiding. The experimental results demonstrate that the EDH is competitive concerning execution time and outperforms the existing masking heuristic algorithms in terms of side effects by reducing information loss of non-sensitive patterns significantly and not creating fake patterns.

Behavior Based Group Recommendation from Social Media Dataset by Using Deep Learning and Topic Modeling

In this digital era, users frequently share their thoughts, preferences, and ideas through social media, which reflect their Basic Human Values. Basic Human Values (aka values) are the fundamental aspects of human behavior, which define what we consider important, and worth having and pursue them. Existing studies identify the values of individuals from different social network usages such as Facebook and Reddit. However, discovering the similarity (or diversity) of value priorities among the members in a group is important since we can reveal many interesting insights such as finding a set of target customers, identifying the chain of misdeed groups, searching for similar acquaintances in workplaces, etc. In this paper, a graph dataset is compiled using the strongest correlation among the features and then we apply a graph clustering technique to identify a suitable hedonist group (i.e., one dimension of values) for users’ recommendations. Then, we also propose a behavior based (i.e., value ) group recommendation technique by analyzing users’ contextual and psychological attributes. Finally, we validate those group members in real life by introducing two hypotheses. In particular, we analyze the tweets of a total of 1140 users collected from Twitter. We obtain a substantial intra-cluster correlation coefficient (ICC) and silhouette clustering coefficient (SCC) scores of 65% and 76%, respectively, among the members in our discovered group.

Enhanced Data Mining and Visualization of Sensory-Graph-Modeled Datasets through Summarization.

The acquisition, processing, mining, and visualization of sensory data for knowledge discovery and decision support has recently been a popular area of research and exploration. Its usefulness is paramount because of its relationship to the continuous involvement in the improvement of healthcare and other related disciplines. As a result of this, a huge amount of data have been collected and analyzed. These data are made available for the research community in various shapes and formats; their representation and study in the form of graphs or networks is also an area of research which many scholars are focused on. However, the large size of such graph datasets poses challenges in data mining and visualization. For example, knowledge discovery from the Bio-Mouse-Gene dataset, which has over 43 thousand nodes and 14.5 million edges, is a non-trivial job. In this regard, summarizing the large graphs provided is a useful alternative. Graph summarization aims to provide the efficient analysis of such complex and large-sized data; hence, it is a beneficial approach. During summarization, all the nodes that have similar structural properties are merged together. In doing so, traditional methods often overlook the importance of personalizing the summary, which would be helpful in highlighting certain targeted nodes. Personalized or context-specific scenarios require a more tailored approach for accurately capturing distinct patterns and trends. Hence, the concept of personalized graph summarization aims to acquire a concise depiction of the graph, emphasizing connections that are closer in proximity to a specific set of given target nodes. In this paper, we present a faster algorithm for the personalized graph summarization (PGS) problem, named IPGS; this has been designed to facilitate enhanced and effective data mining and visualization of datasets from various domains, including biosensors. Our objective is to obtain a similar compression ratio as the one provided by the state-of-the-art PGS algorithm, but in a faster manner. To achieve this, we improve the execution time of the current state-of-the-art approach by using weighted, locality-sensitive hashing, through experiments on eight large publicly available datasets. The experiments demonstrate the effectiveness and scalability of IPGS while providing a similar compression ratio to the state-of-the-art approach. In this way, our research contributes to the study and analysis of sensory datasets through the perspective of graph summarization. We have also presented a detailed study on the Bio-Mouse-Gene dataset, which was conducted to investigate the effectiveness of graph summarization in the domain of biosensors.

A semantic backdoor attack against graph convolutional networks

Graph convolutional networks (GCNs) have been very effective in addressing the issue of various graph-structured related tasks, such as node classification and graph classification. However, recent research has shown that GCNs are vulnerable to a new type of threat called a backdoor attack, where the adversary can inject a hidden backdoor into GCNs so that the attacked model performs well on benign samples, but its prediction will be maliciously changed to the attacker-specified target class if the hidden backdoor is activated by the attacker-defined trigger. A semantic backdoor attack is a new type of backdoor attack on deep neural networks (DNNs), where a naturally occurring semantic feature of samples can serve as a backdoor trigger such that the infected DNNs models will misclassify testing samples containing the predefined semantic feature even without the requirement of modifying the testing samples. Since the backdoor trigger is a naturally occurring semantic feature of the samples, semantic backdoor attacks are more imperceptible and pose a new and serious threat. Existing research on semantic backdoor attacks focuses on the tasks of CNNs-based (Convolutional Neural Networks) image classification and LSTM-based (Long Short-Term Memory) text classification or word prediction. Little attention has been given to semantic backdoor attacks on GCNs models.In this paper, we investigate whether such semantic backdoor attacks are possible for GCNs and propose a semantic backdoor attack against GCNs (SBAG) under the context of graph classification to reveal the existence of this security vulnerability in GCNs. SBAG uses a certain type of nodes in the samples as a backdoor trigger and injects a hidden backdoor into GCNs models by poisoning training data. The backdoor will be activated, and the GCNs models will give malicious classification results specified by the attacker even on unmodified samples as long as the samples contain enough trigger nodes. We evaluate SBAG on five graph datasets. The experimental results indicate that SBAG can achieve attack success rates of approximately 99.9% on unmodified testing samples that naturally contain the trigger and attack success rates over 82% on testing samples modified to inject the trigger, respectively, both under poisoning rates of less than 5%. 11Our code is available at: github.

Attributed graph subspace clustering with residual compensation guided by adaptive dual manifold regularization

As an important technology for recommendation system, attributed graph clustering has received extensive attention recently. Attributed graph clustering methods based on graph neural networks are the mainstream methods . However, their assumption that the attributes of adjacent nodes are similar is often not satisfied in the real world, which influences the clustering performance. To this end, this paper proposes an attributed graph subspace clustering algorithm with residual compensation guided by adaptive dual manifold regularization (ADMRGC). On the basis of the low rank representation subspace clustering (LRR) model, ADMRGC introduces attributed manifold regularization, topological manifold regularization and residual compensation, which make ADMRGC possible to utilize attributed similarity and topological similarity at the same time, thus solving the problem that traditional subspace clustering only considers attributed information. In addition, ADMRGC balances the contribution of node attributes and topology by adaptively weighting the dual manifold regularization, and uses the residual representation matrix to describe the difference between node attributed similarity and topological neighbor relationship. Therefore, ADMRGC can avoid the limitation of graph neural network models assuming that the attributes of adjacent nodes are similar. Experimental results on 7 public graph datasets show that ADMRGC can achieve the best clustering performance on both high-homophily and low-homophily datasets. Especially, for datasets with low homophily, ADMRGC as a shallow model can also achieve better clustering performance than state-of-the-art deep neural network models.