Discriminative Graph-level Anomaly Detection via Dual-students-teacher Model

TL;DR

This paper proposes a novel dual-students-teacher framework for graph-level anomaly detection, effectively distinguishing abnormal graphs by leveraging divergent graph representations and representation errors.

Contribution

It introduces a discriminative graph-level anomaly detection method using a dual-students-teacher model, emphasizing the importance of an effective anomaly score function.

Findings

Effective detection of anomalous graphs demonstrated on real-world datasets.

The dual-students-teacher model outperforms existing methods in accuracy.

Representation errors effectively discriminate abnormal graphs.

Abstract

Different from the current node-level anomaly detection task, the goal of graph-level anomaly detection is to find abnormal graphs that significantly differ from others in a graph set. Due to the scarcity of research on the work of graph-level anomaly detection, the detailed description of graph-level anomaly is insufficient. Furthermore, existing works focus on capturing anomalous graph information to learn better graph representations, but they ignore the importance of an effective anomaly score function for evaluating abnormal graphs. Thus, in this work, we first define anomalous graph information including node and graph property anomalies in a graph set and adopt node-level and graph-level information differences to identify them, respectively. Then, we introduce a discriminative graph-level anomaly detection framework with dual-students-teacher model, where the teacher model with a heuristic loss are trained to make graph representations more divergent. Then, two competing student models trained by normal and abnormal graphs respectively fit graph representations of the teacher model in terms of node-level and graph-level representation perspectives. Finally, we combine representation errors between two student models to discriminatively distinguish anomalous graphs. Extensive experiment analysis demonstrates that our method is effective for the graph-level anomaly detection task on graph datasets in the real world.