APAN: Asynchronous Propagation Attention Network for Real-time Temporal Graph Embedding

TL;DR

This paper introduces APAN, an asynchronous dynamic graph embedding method that decouples graph computation from inference, enabling real-time processing with significantly improved speed for applications like fraud detection.

Contribution

The paper presents a novel asynchronous propagation attention network that separates graph querying from inference, achieving faster real-time embedding in dynamic graphs.

Findings

Achieves 8.7x faster inference speed.

Maintains competitive embedding performance.

Enables real-time applications like fraud detection.

Abstract

Limited by the time complexity of querying k-hop neighbors in a graph database, most graph algorithms cannot be deployed online and execute millisecond-level inference. This problem dramatically limits the potential of applying graph algorithms in certain areas, such as financial fraud detection. Therefore, we propose Asynchronous Propagation Attention Network, an asynchronous continuous time dynamic graph algorithm for real-time temporal graph embedding. Traditional graph models usually execute two serial operations: first graph computation and then model inference. We decouple model inference and graph computation step so that the heavy graph query operations will not damage the speed of model inference. Extensive experiments demonstrate that the proposed method can achieve competitive performance and 8.7 times inference speed improvement in the meantime.