Adaptive Graph Encoder for Attributed Graph Embedding

Cui, Guixiang; Zhou, Jie; Yang, Cheng; Liu, Zhiyuan

doi:10.1145/3394486.3403140

Cited by 161 publications

(103 citation statements)

References 23 publications

(20 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Adaptive Graph Encoder (AGE) is another filter-related graph embedding model for community detection. Cui et al [42] proposed a two-part model for auto-encoder based community detection. The model consists of a Laplacian smoothing filter and an adaptive encoder.…”

Section: Ae-based Community Detection Strategiesmentioning

confidence: 99%

A Survey on the Recent Advances of Deep Community Detection

2021

View full text Add to dashboard Cite

In the first days of social networking, the typical view of a community was a set of user profiles of the same interests and likes, and this community kept enlarging by searching, proposing, and adding new members with the same characteristics that were likely to interfere with the existing members. Today, things have changed dramatically. Social networking platforms are not restricted to forming similar user profiles: The vast amounts of data produced every day have given opportunities to predict and suggest relationships, behaviors, and everyday activities like shopping, food, traveling destinations, etc. Every day, vast data amounts are generated by the famous social networks such as Facebook, Twitter, Instagram, and so on. For example, Facebook alone generates 4 petabytes of data per day. The analysis of such data is of high importance to many aspects like recommendation systems, businesses, health organizations, etc. The community detection problem received considerable attention a long time ago. Communities are represented as clusters of an entire network. Most of the community detection techniques are based on graph structures. In this paper, we present the recent advances of deep learning techniques for community detection. We describe the most recent strategies presented in this field, and we provide some general discussion and some future trends.

show abstract

Section: Ae-based Community Detection Strategiesmentioning

confidence: 99%

A Survey on the Recent Advances of Deep Community Detection

2021

View full text Add to dashboard Cite

show abstract

“…On the other hand, proposing a perturbation filtering mechanism to reduce the size of multinode candidate perturbations set is also an effective way. In addition, our method does not consider the constraints of attributed graphs [33], such as attribution-based node similarity constraint [34] and attribution cooccurrence constraint [17]. Parallel multinode adversarial attack on attributed graph and Heterogeneous Information Network (HIN) [35] still needs further exploration.…”

Section: Adversarial Attack On Graphsmentioning

confidence: 99%

Towards Revealing Parallel Adversarial Attack on Politician Socialnet of Graph Structure

Tian

Liu

Tong

et al. 2021

Security and Communication Networks

View full text Add to dashboard Cite

Socialnet becomes an important component in real life, drawing a lot of study issues of security and safety. Recently, for the features of graph structure in socialnet, adversarial attacks on node classification are exposed, and automatic attack methods such as fast gradient attack (FGA) and NETTACK are developed for per-node attacks, which can be utilized for multinode attacks in a sequential way. However, due to the overlook of perturbation influence between different per-node attacks, the above sequential method does not guarantee a global attack success rate for all target nodes, under a fixed budget of perturbation. In this paper, we propose a parallel adversarial attack framework on node classification. We redesign new loss function and objective function for nonconstraint and constraint perturbations, respectively. Through constructing intersection and supplement mechanisms of perturbations, we then integrate node filtering-based P-FGA and P-NETTACK in a unified framework, finally realizing parallel adversarial attacks. Experiments on politician socialnet dataset Polblogs with detailed analysis are conducted to show the effectiveness of our approach.

show abstract

“…• Node classification: GCN [11], GAT [24], simplified GCN (SGC) [28], GraphSAGE [8], JK-Net [30], APPNP [13], AP-GCN [22],FastGCN [4], Scalable Inception Graph Neural Networks (SIGN) [20], ClusterGCN [5] and GraphSAINT [32]. • Link prediction: GAE and VGAE [12], LightGCN [10], simplified GCN (SGC) [28], Scalable Inception Graph Neural Networks (SIGN) [20], Adaptive Graph Encoder (AGE) [6], Spectral Clustering (SC) [17] and DeepWalk (DW) [19]. • Node clustering: GAE and VGAE [12], MGAE [26], ARGA and ARVGA [18], AGC [36], DAEGC [25], SDCN [2] and AGE [6].…”

Section: Datasets and Baselinesmentioning

confidence: 99%

“…• Link prediction: GAE and VGAE [12], LightGCN [10], simplified GCN (SGC) [28], Scalable Inception Graph Neural Networks (SIGN) [20], Adaptive Graph Encoder (AGE) [6], Spectral Clustering (SC) [17] and DeepWalk (DW) [19]. • Node clustering: GAE and VGAE [12], MGAE [26], ARGA and ARVGA [18], AGC [36], DAEGC [25], SDCN [2] and AGE [6]. A detailed introduction of these baseline methods can be found in Section A.2 of the Appendix.…”

Section: Datasets and Baselinesmentioning

confidence: 99%

Rod

Zhang

Jiang

et al. 2021

Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

Graph neural networks (GNNs) have been widely used in many graph-based tasks such as node classification, link prediction, and node clustering. However, GNNs gain their performance benefits mainly from performing the feature propagation and smoothing across the edges of the graph, thus requiring sufficient connectivity and label information for effective propagation. Unfortunately, many real-world networks are sparse in terms of both edges and labels, leading to sub-optimal performance of GNNs. Recent interest in this sparse problem has focused on the self-training approach, which expands supervised signals with pseudo labels. Nevertheless, the self-training approach inherently cannot realize the full potential of refining the learning performance on sparse graphs due to the unsatisfactory quality and quantity of pseudo labels.In this paper, we propose ROD, a novel reception-aware online knowledge distillation approach for sparse graph learning. We design three supervision signals for ROD: multi-scale reception-aware graph knowledge, task-based supervision, and rich distilled knowledge, allowing online knowledge transfer in a peer-teaching manner. To extract knowledge concealed in the multi-scale reception fields, ROD explicitly requires individual student models to preserve different levels of locality information. For a given task, each student would predict based on its reception-scale knowledge, while simultaneously a strong teacher is established on-the-fly by combining multi-scale knowledge. Our approach has been extensively evaluated on 9 datasets and a variety of graph-based tasks, including node classification, link prediction, and node clustering. The result demonstrates that ROD achieves state-of-art performance and is more robust for the graph sparsity. CCS CONCEPTS• Mathematics of computing → Graph algorithms; • Computing methodologies → Neural networks.

show abstract

Adaptive Graph Encoder for Attributed Graph Embedding

Cited by 161 publications

References 23 publications

A Survey on the Recent Advances of Deep Community Detection

A Survey on the Recent Advances of Deep Community Detection

Towards Revealing Parallel Adversarial Attack on Politician Socialnet of Graph Structure

Rod

Contact Info

Product

Resources

About