Long Range Propagation on Continuous-Time Dynamic Graphs

TL;DR

This paper introduces CTAN, a novel continuous-time graph neural network designed to effectively model long-range dependencies in dynamic graphs, outperforming existing methods on synthetic and real-world benchmarks.

Contribution

The paper proposes CTAN, a continuous-time graph neural network that efficiently captures long-range dependencies in dynamic graphs, supported by theoretical analysis and empirical results.

Findings

CTAN outperforms existing methods on synthetic long-range benchmarks.

CTAN demonstrates superior performance on real-world dynamic graph tasks.

Theoretical analysis confirms CTAN's capability for long-range information propagation.

Abstract

Learning Continuous-Time Dynamic Graphs (C-TDGs) requires accurately modeling spatio-temporal information on streams of irregularly sampled events. While many methods have been proposed recently, we find that most message passing-, recurrent- or self-attention-based methods perform poorly on long-range tasks. These tasks require correlating information that occurred "far" away from the current event, either spatially (higher-order node information) or along the time dimension (events occurred in the past). To address long-range dependencies, we introduce Continuous-Time Graph Anti-Symmetric Network (CTAN). Grounded within the ordinary differential equations framework, our method is designed for efficient propagation of information. In this paper, we show how CTAN's (i) long-range modeling capabilities are substantiated by theoretical findings and how (ii) its empirical performance on synthetic long-range benchmarks and real-world benchmarks is superior to other methods. Our results motivate CTAN's ability to propagate long-range information in C-TDGs as well as the inclusion of long-range tasks as part of temporal graph models evaluation.