Cross-Domain Graph Level Anomaly Detection

doi:10.1109/TKDE.2024.3462442

Cross-Domain Graph Level Anomaly Detection

2024-12-01

0

OA

AI

DOI：10.1109/TKDE.2024.3462442

Original

Original request for help

Save

Abstract

En 中文

Existing graph level anomaly detection methods are predominantly unsupervised due to high costs for obtaining labels, yielding sub-optimal detection accuracy when compared to supervised methods. Moreover, they heavily rely on the assumption that the training data exclusively consists of normal graphs. Hence, even the presence of a few anomalous graphs can lead to substantial performance degradation. To alleviate these problems, we propose a cross-domain graph level anomaly detection method, aiming to identify anomalous graphs from a set of unlabeled graphs (target domain) by using easily accessible normal graphs from a different but related domain (source domain). Our method consists of four components: a feature extractor that preserves semantic and topological information of individual graphs while incorporating the distance between different graphs; an adversarial domain classifier to make graph level representations domain-invariant; a one-class classifier to exploit label information in the source domain; and a class aligner to align classes from both domains based on pseudolabels. Experiments on seven benchmark datasets show that the proposed method largely outperforms state-of-the-art methods.

Keywords:

Anomaly detection

Feature extraction

Semantics

Databases

Training data

Training

Transfer learning

Graph anomaly detection

graph neural networks

graph transfer learning

Cross-Domain Graph Level Anomaly Detection

Abstract

Journal

Organization