GraphAnoGAN: Detecting Anomalous Snapshots from Attributed Graphs

TL;DR

GraphAnoGAN introduces a novel framework combining generative and discriminative models to effectively detect anomalous snapshots in attributed graphs, outperforming existing shallow learning methods by leveraging structure-attribute interactions.

Contribution

The paper presents GraphAnoGAN, a new deep learning framework that models complex interactions in attributed graphs for anomaly detection, surpassing prior shallow approaches.

Findings

Outperforms 6 baseline methods with 28.29% higher precision

Achieves 22.01% higher recall on average

Effective on 4 real-world network datasets

Abstract

Finding anomalous snapshots from a graph has garnered huge attention recently. Existing studies address the problem using shallow learning mechanisms such as subspace selection, ego-network, or community analysis. These models do not take into account the multifaceted interactions between the structure and attributes in the network. In this paper, we propose GraphAnoGAN, an anomalous snapshot ranking framework, which consists of two core components -- generative and discriminative models. Specifically, the generative model learns to approximate the distribution of anomalous samples from the candidate set of graph snapshots, and the discriminative model detects whether the sampled snapshot is from the ground-truth or not. Experiments on 4 real-world networks show that GraphAnoGAN outperforms 6 baselines with a significant margin (28.29% and 22.01% higher precision and recall, respectively compared to the best baseline, averaged across all datasets).