A Comprehensive Survey on Graph Anomaly Detection With Deep Learning (2021-06-14T00:00:00.000000Z)

TL;DR

A taxonomy that follows a task-driven strategy and categorizes existing work according to the anomalous graph objects that they can detect is provided, to create a “one-stop-shop” that provides a unified understanding of the problem categories and existing approaches.

Abstract

Anomalies are rare observations (e.g., data records or events) that deviate significantly from the others in the sample. Over the past few decades, research on anomaly mining has received increasing interests due to the implications of these occurrences in a wide range of disciplines - for instance, security, finance, and medicine. For this reason, anomaly detection, which aims to identify these rare observations, has become one of the most vital tasks in the world and has shown its power in preventing detrimental events, such as financial fraud, network intrusions, and social spam. The detection task is typically solved by identifying outlying data points in the feature space, which, inherently, overlooks the relational information in real-world data. At the same time, graphs have been prevalently used to represent the structural/relational information, which raises the graph anomaly detection problem - identifying anomalous graph objects (i.e., nodes, edges and sub-graphs) in a single graph, or anomalous graphs in a set/database of graphs. Conventional anomaly detection techniques cannot tackle this problem well because of the complexity of graph data (e.g., irregular structures, relational dependencies, node/edge types/attributes/directions/multiplicities/weights, large scale, etc.). However, thanks to the advent of deep learning in breaking these limitations, graph anomaly detection with deep learning has received a growing attention recently. In this survey, we aim to provide a systematic and comprehensive review of the contemporary deep learning techniques for graph anomaly detection. Specifically, we provide a taxonomy that follows a task-driven strategy and categorizes existing work according to the anomalous graph objects that they can detect. We especially focus on the challenges in this research area and discuss the key intuitions, technical details as well as relative strengths and weaknesses of various techniques in each category. From the survey results, we highlight 12 future research directions spanning unsolved and emerging problems introduced by graph data, anomaly detection, deep learning and real-world applications. Additionally, to provide a wealth of useful resources for future studies, we have compiled a set of open-source implementations, public datasets, and commonly-used evaluation metrics. With this survey, our goal is to create a “one-stop-shop” that provides a unified understanding of the problem categories and existing approaches, publicly available hands-on resources, and high-impact open challenges for graph anomaly detection using deep learning.

References225 items

eFraudCom: An E-commerce Fraud Detection System via Competitive Graph Neural Networks

ComGA: Community-Aware Attributed Graph Anomaly Detection

Combining Machine Learning with Knowledge Engineering to detect Fake News in Social Networks - A Survey

FRAUDRE: Fraud Detection Dual-Resistant to Graph Inconsistency and Imbalance

Anomaly Mining: Past, Present and Future

Cross-Domain Graph Anomaly Detection

Decoupling Representation Learning and Classification for GNN-based Anomaly Detection

Outlier Detection in High Dimensional Data

A Comprehensive Survey on Community Detection With Deep Learning

Anomaly Attribution with Likelihood Compensation

User Preference-aware Fake News Detection

Anomaly Detection on Attributed Networks via Contrastive Self-Supervised Learning

Few-shot Network Anomaly Detection via Cross-network Meta-learning

Local distribution-based adaptive minority oversampling for imbalanced data classification

SUGAR: Subgraph Neural Network with Reinforcement Pooling and Self-Supervised Mutual Information Mechanism

On Using Classification Datasets to Evaluate Graph Outlier Detection: Peculiar Observations and New Insights

F-FADE: Frequency Factorization for Anomaly Detection in Edge Streams

AANE: Anomaly Aware Network Embedding For Anomalous Link Detection

AtNE-Trust: Attributed Trust Network Embedding for Trust Prediction in Online Social Networks

Exploratory Adversarial Attacks on Graph Neural Networks

Graph Geometry Interaction Learning

Heterogeneous Hypergraph Embedding for Graph Classification

Error-Bounded Graph Anomaly Loss for GNNs

Approximations

SwissLog: Robust and Unified Deep Learning Based Log Anomaly Detection for Diverse Faults

ResGCN: attention-based deep residual modeling for anomaly detection on attributed networks

A Unifying Review of Deep and Shallow Anomaly Detection

Automating Outlier Detection via Meta-Learning

MStream: Fast Anomaly Detection in Multi-Aspect Streams

Attack signal estimation for intrusion detection in industrial control system

Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters

FANG: Leveraging Social Context for Fake News Detection Using Graph Representation

Deep spatial–temporal structure learning for rumor detection on Twitter

A Deep Multi-View Framework for Anomaly Detection on Attributed Networks

Boosting label weighted extreme learning machine for classifying multi-label imbalanced data

DATE: Dual Attentive Tree-aware Embedding for Customs Fraud Detection

Fraud Transactions Detection via Behavior Tree with Local Intention Calibration

Ultrafast Local Outlier Detection from a Data Stream with Stationary Region Skipping

A comprehensive survey of anomaly detection techniques for high dimensional big data

Unified Graph Embedding-Based Anomalous Edge Detection

Towards a Hierarchical Bayesian Model of Multi-View Anomaly Detection

Multi-View Attribute Graph Convolution Networks for Clustering

Inductive Anomaly Detection on Attributed Networks

Backdoor Attacks to Graph Neural Networks

Subgraph Neural Networks

Heterogeneous Graph Attention Networks for Early Detection of Rumors on Twitter

Multi-level Graph Convolutional Networks for Cross-platform Anchor Link Prediction

Evaluation of Machine Learning Algorithms for Anomaly Detection

Fraud detection: A systematic literature review of graph-based anomaly detection approaches

Toward a Better Performance Evaluation Framework for Fake News Classification

GCN-Based User Representation Learning for Unifying Robust Recommendation and Fraudster Detection

Deep Learning for Community Detection: Progress, Challenges and Opportunities

Structural Temporal Graph Neural Networks for Anomaly Detection in Dynamic Graphs

Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection

GCAN: Graph-aware Co-Attention Networks for Explainable Fake News Detection on Social Media

One2Multi Graph Autoencoder for Multi-view Graph Clustering

Beyond Rank-1: Discovering Rich Community Structure in Multi-Aspect Graphs

MixedAD: A Scalable Algorithm for Detecting Mixed Anomalies in Attributed Graphs

Anomalous Instance Detection in Deep Learning: A Survey

Self-Trained Deep Ordinal Regression for End-to-End Video Anomaly Detection

One-class graph neural networks for anomaly detection in attributed networks

OCGNN: One-class Classification with Graph Neural Networks

Interleaved Sequence RNNs for Fraud Detection

Anomalydae: Dual Autoencoder for Anomaly Detection on Attributed Networks

MAGNN: Metapath Aggregated Graph Neural Network for Heterogeneous Graph Embedding

Machine Learning Techniques for Network Anomaly Detection: A Survey

Outlier Resistant Unsupervised Deep Architectures for Attributed Network Embedding

DySAT: Deep Neural Representation Learning on Dynamic Graphs via Self-Attention Networks

Multi-scale Anomaly Detection on Attributed Networks

MIDAS: Microcluster-Based Detector of Anomalies in Edge Streams

A Semi-Supervised Graph Attentive Network for Financial Fraud Detection

Block-Structured Optimization for Anomalous Pattern Detection in Interdependent Networks

Selective network discovery via deep reinforcement learning on embedded spaces

Jointly Embedding the Local and Global Relations of Heterogeneous Graph for Rumor Detection

Salient Subsequence Learning for Time Series Clustering

A hierarchical contextual attention-based network for sequential recommendation

SliceNDice: Mining Suspicious Multi-Attribute Entity Groups with Multi-View Graphs

SpecAE: Spectral AutoEncoder for Anomaly Detection in Attributed Networks

Semi-Supervised Learning and Graph Neural Networks for Fake News Detection

AddGraph: Anomaly Detection in Dynamic Graph Using Attention-based Temporal GCN

Heterogeneous Graph Neural Network

Deep Anomaly Detection with Deviation Networks

Fast and Accurate Anomaly Detection in Dynamic Graphs with a Two-Pronged Approach

Multi-View Anomaly Detection: Neighborhood in Locality Matters

Deep Structured Cross-Modal Anomaly Detection

A Robust Embedding Method for Anomaly Detection on Attributed Networks

Deep Semi-Supervised Anomaly Detection

Spotting Collective Behaviour of Online Frauds in Customer Reviews

Representation Learning for Dynamic Graphs: A Survey

Securing the Deep Fraud Detector in Large-Scale E-Commerce Platform via Adversarial Machine Learning Approach

FdGars: Fraudster Detection via Graph Convolutional Networks in Online App Review System

Deep Anomaly Detection on Attributed Networks

Time series feature learning with labeled and unlabeled data

From Anomaly Detection to Rumour Detection using Data Streams of Social Platforms

Summit: Scaling Deep Learning Interpretability by Visualizing Activation and Attribution Summarizations

An unsupervised parameter learning model for RVFL neural network

Anomaly Detection for an E-commerce Pricing System

Multi-GCN: Graph Convolutional Networks for Multi-View Networks, with Applications to Global Poverty

Interactive Anomaly Detection on Attributed Networks

100

Detecting and Assessing Anomalous Evolutionary Behaviors of Nodes in Evolving Social Networks

101

Anomaly Detection in the Dynamics of Web and Social Networks Using Associative Memory

102

Deep Learning for Anomaly Detection

103

Deep Learning for Anomaly Detection: A Survey

104

Graph Neural Networks: A Review of Methods and Applications

105

ICANE: interaction content-aware network embedding via co-embedding of nodes and edges

106

Outlier Aware Network Embedding for Attributed Networks

107

A Hierarchical Multi-task Approach for Learning Embeddings from Semantic Tasks

108

Deep Structure Learning for Fraud Detection

109

SedanSpot: Detecting Anomalies in Edge Streams

110

Heterogeneous Graph Neural Networks for Malicious Account Detection

111

ChangeDAR: Online Localized Change Detection for Sensor Data on a Graph

112

Multi-task Deep Reinforcement Learning with PopArt

113

NetWalk: A Flexible Deep Embedding Approach for Anomaly Detection in Dynamic Networks

114

SpotLight: Detecting Anomalies in Streaming Graphs

115

Deep One-Class Classification

116

A comprehensive survey on network anomaly detection

117

Multiple Structure-View Learning for Graph Classification

118

ANOMALOUS: A Joint Modeling Approach for Anomaly Detection on Attributed Networks

119

Deep into Hypersphere: Robust and Unsupervised Anomaly Discovery in Dynamic Networks

120

FraudNE: a Joint Embedding Approach for Fraud Detection

121

Learning Representations of Ultrahigh-dimensional Data for Random Distance-based Outlier Detection

122

Adversarial Attack on Graph Structured Data

123

Bayesian Robust Attributed Graph Clustering: Joint Learning of Partial Anomalies and Group Structure

124

An Efficient Framework for Detecting Evolving Anomalous Subgraphs in Dynamic Networks

125

Group Anomaly Detection using Deep Generative Models

126

One-Class Adversarial Nets for Fraud Detection

127

Deep Autoencoding Gaussian Mixture Model for Unsupervised Anomaly Detection

128

REV2: Fraudulent User Prediction in Rating Platforms

129

Fast, Accurate, and Flexible Algorithms for Dense Subtensor Mining

130

Visual interpretability for deep learning: a survey

131

A Survey on Network Embedding

132

Anomaly Detection in Dynamic Networks using Multi-view Time-Series Hypersphere Learning

133

GANG: Detecting Fraudulent Users in Online Social Networks via Guilt-by-Association on Directed Graphs

134

Graph Attention Networks

135

A survey of deep learning-based network anomaly detection

136

Representation Learning on Graphs: Methods and Applications

137

Radar: Residual Analysis for Anomaly Detection in Attributed Networks

138

Anomaly Detection with Robust Deep Autoencoders

139

Accelerated Local Anomaly Detection via Resolving Attributed Networks

140

graph2vec: Learning Distributed Representations of Graphs

141

Graph-Based Fraud Detection in the Face of Camouflage

142

DenseAlert: Incremental Dense-Subtensor Detection in Tensor Streams

143

Inductive Representation Learning on Large Graphs

144

Detecting users’ anomalous emotion using social media for business intelligence☆

145

Adaptive Spammer Detection with Sparse Group Modeling

146

Good Semi-supervised Learning That Requires a Bad GAN

147

Malevolent Activity Detection with Hypergraph-Based Models

148

SEANO: Semi-supervised Embedding in Attributed Networks with Outliers

149

Collaborative Dynamic Sparse Topic Regression with User Profile Evolution for Item Recommendation

150

Variational Graph Auto-Encoders

151

High-dimensional and large-scale anomaly detection using a linear one-class SVM with deep learning

152

Semi-Supervised Classification with Graph Convolutional Networks

153

FRAUDAR: Bounding Graph Fraud in the Face of Camouflage

154

node2vec: Scalable Feature Learning for Networks

155

A Scalable Approach for Outlier Detection in Edge Streams Using Sketch-based Approximations

156

An embedding approach to anomaly detection

157

Fast Memory-efficient Anomaly Detection in Streaming Heterogeneous Graphs

158

Scalable Anomaly Ranking of Attributed Neighborhoods

159

A Survey on Social Media Anomaly Detection

160

BIRDNEST: Bayesian Inference for Ratings-Fraud Detection

161

EdgeCentric: Anomaly Detection in Edge-Attributed Networks

162

Temporal Multi-View Inconsistency Detection for Network Traffic Analysis

163

Anomaly detection in dynamic networks: a survey

164

LINE: Large-scale Information Network Embedding

165

Boosting for Multi-Graph Classification

166

Unsupervised Learning of Video Representations using LSTMs

167

Less is More: Building Selective Anomaly Ensembles

168

Multi-graph-view Learning for Graph Classification

169

Spotting Suspicious Link Behavior with fBox: An Adversarial Perspective

170

A unified approach to network anomaly detection

171

Anomaly detection in online social networks

172

Focused clustering and outlier detection in large attributed graphs

173

Graph based anomaly detection and description: a survey

174

DeepWalk: online learning of social representations

175

Bag Constrained Structure Pattern Mining for Multi-Graph Classification

176

Near-optimal Anomaly Detection in Graphs using Lovasz Extended Scan Statistic

177

Multi-instance Multi-graph Dual Embedding Learning

178

Statistical Selection of Congruent Subspaces for Mining Attributed Graphs

179

Efficient anomaly detection in dynamic, attributed graphs: Emerging phenomena and big data

180

Modeling dynamic behavior in large evolving graphs

181

Intrusion as (anti)social communication: characterization and detection

182

Community-based anomaly detection in evolutionary networks

183

Temporal link prediction by integrating content and structure information

184

Outlier detection in graph streams

185

A Survey of Outlier Detection Methods in Network Anomaly Identification

186

oddball: Spotting Anomalies in Weighted Graphs

187

Streaming k-means approximation

188

RTG: a recursive realistic graph generator using random typing

189

Community detection algorithms: a comparative analysis: invited presentation, extended abstract

190

Anomaly detection: A survey

191

Hypergraph-Based Anomaly Detection of High-Dimensional Co-Occurrences

192

CUR matrix decompositions for improved data analysis

193

Spotting Significant Changing Subgraphs in Evolving Graphs

194

Collective Classification in Network Data

195

Outrank: a Graph-Based Outlier Detection Framework Using Random Walk

196

Netprobe: a fast and scalable system for fraud detection in online auction networks

197

Detection of emerging space-time clusters

198

Graph-based anomaly detection

199

LOF: identifying density-based local outliers

200

A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs

201

Collective dynamics of ‘small-world’ networks

202

A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise

203

How to Detect and Handle Outliers

204

Procedures for Detecting Outlying Observations in Samples

205

GLAD-PAW: Graph-Based Log Anomaly Detection by Position Aware Weighted Graph Attention Network

206

detec-JOURNAL OF L A TEX CLASS FILES, VOL. , NO. , AUGUST 2021 tion

207

Anomaly Detection for Big Data Using Efficient Techniques: A Review

208

Evaluation of Anomaly-Based Intrusion Detection with Combined Imbalance Correction and Feature Selection

209

Graph Stochastic Neural Networks for Semi-supervised Learning

210

Evaluating Attribution for Graph Neural Networks

211

A Comprehensive Survey on Graph Neural Networks

212

GENERATIVE ADVERSARIAL NETS

213

Hunt For The Unique, Stable, Sparse And Fast Feature Learning On Graphs

214

Node Classification in Signed Social Networks

215

Identifying Connectivity Patterns for Brain Diseases via Multi-side-view Guided Deep Architectures

216

Spatial Scan Statistic

217

Multi-Graph Learning with Positive and Unlabeled Bags

218

A Probabilistic Approach to Uncovering Attributed Graph Anomalies

219

NetSpot: Spotting Significant Anomalous Regions on Dynamic Networks

220

Collective Classi!cation in Network Data

221

The Epoch-Greedy Algorithm for Multi-armed Bandits with Side Information

222

Goodness-of-fit test statistics that dominate the Kolmogorov statistics

223

Deﬁnition 2

224

and conference papers

225