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

Peng, Zhen; Luo, Minnan; Li, Jundong; Xue, Luguo; Zheng, Qinghua

doi:10.1109/tkde.2020.3015098

Cited by 40 publications

(21 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, since the GCN has structural limitations in the interpretation of complex and dynamic relationships in graphs, some anomaly detection studies that hybridize GCN and GAT using the concept of attention have been proposed [52,53]. Since the GAD for the attributes network assumes one complex graph in which graphs of various domains are mixed, they focus on finding a dimension that can interpret the complex dimension of the graph by hybridizing various models instead of using one GCN model [54]. In other words, it aims to train a comprehensive and generalized model that can detect various types of anomalies from various data sources and graphs.…”

Section: Graph-based Anomaly Detectionmentioning

confidence: 99%

Graph Representation Learning-Based Early Depression Detection Framework in Smart Home Environments

Kim

Sohn

2022

Sensors

View full text Add to dashboard Cite

Although the diagnosis and treatment of depression is a medical field, ICTs and AI technologies are used widely to detect depression earlier in the elderly. These technologies are used to identify behavioral changes in the physical world or sentiment changes in cyberspace, known as symptoms of depression. However, although sentiment and physical changes, which are signs of depression in the elderly, are usually revealed simultaneously, there is no research on them at the same time. To solve the problem, this paper proposes knowledge graph-based cyber–physical view (CPV)-based activity pattern recognition for the early detection of depression, also known as KARE. In the KARE framework, the knowledge graph (KG) plays key roles in providing cross-domain knowledge as well as resolving issues of grammatical and semantic heterogeneity required in order to integrate cyberspace and the physical world. In addition, it can flexibly express the patterns of different activities for each elderly. To achieve this, the KARE framework implements a set of new machine learning techniques. The first is 1D-CNN for attribute representation in relation to learning to connect the attributes of physical and cyber worlds and the KG. The second is the entity alignment with embedding vectors extracted by the CNN and GNN. The third is a graph extraction method to construct the CPV from KG with the graph representation learning and wrapper-based feature selection in the unsupervised manner. The last one is a method of activity-pattern graph representation based on a Gaussian Mixture Model and KL divergence for training the GAT model to detect depression early. To demonstrate the superiority of the KARE framework, we performed the experiments using real-world datasets with five state-of-the-art models in knowledge graph entity alignment.

show abstract

Section: Graph-based Anomaly Detectionmentioning

confidence: 99%

Graph Representation Learning-Based Early Depression Detection Framework in Smart Home Environments

Kim

Sohn

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Over-smoothing issue → community-specific representation ALARM [24] Over-smoothing issue → GCN with residual-based attention CoLA [30] (2021)…”

Section: A Gnn-based Static Graph Anomaly Detectionmentioning

confidence: 99%

“…In a multi-view attributed network, each node is affiliated with a set of multi-dimensional features (attributes), which can be represented by k distinct feature spaces along with k views. By making use of the multi-view attributes, Peng et al [24] proposed a deep multi-view framework for detection (ALARM) for detecting global and structural anomalies. Specifically, it employs multiple GCNs to embed a multi-view attributed graph as an encoder and uses two decoders, each of which reconstructs the graph structure and attributes.…”

Section: A Gnn-based Static Graph Anomaly Detectionmentioning

confidence: 99%

Graph Anomaly Detection with Graph Neural Networks: Current Status and Challenges

Kim¹,

Lee²,

Shin³

et al. 2022

Preprint

View full text Add to dashboard Cite

Graphs are used widely to model complex systems, and detecting anomalies in a graph is an important task in the analysis of complex systems. Graph anomalies are patterns in a graph that do not conform to normal patterns expected of the attributes and/or structures of the graph. In recent years, graph neural networks (GNNs) have been studied extensively and have successfully performed difficult machine learning tasks in node classification, link prediction, and graph classification thanks to the highly expressive capability via message passing in effectively learning graph representations. To solve the graph anomaly detection problem, GNN-based methods leverage information about the graph attributes (or features) and/or structures to learn to score anomalies appropriately. In this survey, we review the recent advances made in detecting graph anomalies using GNN models. Specifically, we summarize GNN-based methods according to the graph type (i.e., static and dynamic), the anomaly type (i.e., node, edge, subgraph, and whole graph), and the network architecture (e.g., graph autoencoder, graph convolutional network). To the best of our knowledge, this survey is the first comprehensive review of graph anomaly detection methods based on GNNs.

show abstract

“…To enhance the performance of anomalous node detection, later work by Peng et al [90] further explores node attributes from multiple attributed views to detect anomalies. The multiple attributed views are employed to describe different perspectives of the objects [91]- [93].…”

Section: Gcn Based Techniquesmentioning

confidence: 99%

“…In real-world networks, objects might form different kinds of relationships with others (e.g., user's followership and friendship on Twiter) and their attribute information might be collected from different resources (e.g., user's profile, historical posts). This results in two types of multi-view graphs: 1) multi-graph that contains more than one type of edges between two nodes [154], [155]; and 2) multi-attributed-view graph that stores node attributes in different attributed views [90], [156], [157].…”

Section: Multi-view Graph Anomaly Detectionmentioning

confidence: 99%

A Comprehensive Survey on Graph Anomaly Detection with Deep Learning

Ma,

Wu,

Xue

et al. 2021

Preprint

View full text Add to dashboard Cite

Anomalies represent rare observations (e.g., data records, messages or events) that are deviating significantly from others. Over the last forty years, researches on anomalies have received intensified interests and the burst of information has attracted more attention on anomalies because of their significance in a wide range of disciplines (e.g., computer science, chemistry, and biology). Anomaly detection, which aims to identify these rare observations, is among the most vital tasks and has shown its power in preventing detrimental events, such as financial fraud, network intrusion, and social spam, from happening. The detection task is typically solved by detecting outlying data points in the features space and inherently overlooks the structural information in real-world data. Graphs have been prevalently used to preserve the structural information, and this raises the graph anomaly detection problem -identifying anomalous graph objects (i.e., nodes, edges and sub-graphs). However, conventional anomaly detection techniques cannot well solve this problem because of the complexity of graph data (e.g., irregular structures, non-independent and large-scale). For the aptitudes of deep learning in breaking these limitations, graph anomaly detection with deep learning has received intensified studies recently. In this survey, we aim to provide a systematic and comprehensive review of the contemporary deep learning techniques for graph anomaly detection. Specifically, our categorization follows a task-driven strategy and classifies existing works according to the anomalous graph objects they can detect. We especially focus on the motivations, key intuitions and technical details of existing works. Moreover, we summarize open-sourced implementations, public datasets, and commonly-used evaluation metrics to provide affluent resources for future studies. Finally, we highlight twelve extensive future research directions according to our survey results covering emerging problems introduced by graph data, anomaly detection and real applications.

show abstract

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

Cited by 40 publications

References 50 publications

Graph Representation Learning-Based Early Depression Detection Framework in Smart Home Environments

Graph Representation Learning-Based Early Depression Detection Framework in Smart Home Environments

Graph Anomaly Detection with Graph Neural Networks: Current Status and Challenges

A Comprehensive Survey on Graph Anomaly Detection with Deep Learning

Contact Info

Product

Resources

About