SLADE: Detecting Dynamic Anomalies in Edge Streams without Labels via Self-Supervised Learning

TL;DR

SLADE is a self-supervised deep learning method for real-time detection of dynamic anomalies in evolving edge streams, capable of adapting to changing graph states without labeled data, and operates efficiently in constant time per edge.

Contribution

This paper introduces SLADE, a novel self-supervised approach for rapid, label-free anomaly detection in dynamic edge streams that operates efficiently in real-time.

Findings

Outperforms nine competing methods on four real-world datasets

Operates in constant time per new edge in the stream

Effectively detects anomalies without relying on labeled data

Abstract

To detect anomalies in real-world graphs, such as social, email, and financial networks, various approaches have been developed. While they typically assume static input graphs, most real-world graphs grow over time, naturally represented as edge streams. In this context, we aim to achieve three goals: (a) instantly detecting anomalies as they occur, (b) adapting to dynamically changing states, and (c) handling the scarcity of dynamic anomaly labels. In this paper, we propose SLADE (Self-supervised Learning for Anomaly Detection in Edge Streams) for rapid detection of dynamic anomalies in edge streams, without relying on labels. SLADE detects the shifts of nodes into abnormal states by observing deviations in their interaction patterns over time. To this end, it trains a deep neural network to perform two self-supervised tasks: (a) minimizing drift in node representations and (b) generating long-term interaction patterns from short-term ones. Failure in these tasks for a node signals its deviation from the norm. Notably, the neural network and tasks are carefully designed so that all required operations can be performed in constant time (w.r.t. the graph size) in response to each new edge in the input stream. In dynamic anomaly detection across four real-world datasets, SLADE outperforms nine competing methods, even those leveraging label supervision.