Towards OOD Generalization in Dynamic Graphs via Causal Invariant Learning

TL;DR

This paper introduces DyCIL, a causal invariant learning framework for dynamic graphs, addressing out-of-distribution generalization by identifying invariant patterns, capturing evolution rationale, and modeling distributional shifts.

Contribution

The paper proposes a novel DyCIL model that explicitly identifies causal subgraphs, extracts evolution rationale, and models shifts for improved OOD generalization in dynamic graphs.

Findings

DyCIL outperforms baselines on real-world datasets.

Effective identification of causal subgraphs.

Robustness to diverse OOD shifts.

Abstract

Although dynamic graph neural networks (DyGNNs) have demonstrated promising capabilities, most existing methods ignore out-of-distribution (OOD) shifts that commonly exist in dynamic graphs. Dynamic graph OOD generalization is non-trivial due to the following challenges: 1) Identifying invariant and variant patterns amid complex graph evolution, 2) Capturing the intrinsic evolution rationale from these patterns, and 3) Ensuring model generalization across diverse OOD shifts despite limited data distribution observations. Although several attempts have been made to tackle these challenges, none has successfully addressed all three simultaneously, and they face various limitations in complex OOD scenarios. To solve these issues, we propose a Dynamic graph Causal Invariant Learning (DyCIL) model for OOD generalization via exploiting invariant spatio-temporal patterns from a causal view. Specifically, we first develop a dynamic causal subgraph generator to identify causal dynamic subgraphs explicitly. Next, we design a causal-aware spatio-temporal attention module to extract the intrinsic evolution rationale behind invariant patterns. Finally, we further introduce an adaptive environment generator to capture the underlying dynamics of distributional shifts. Extensive experiments on both real-world and synthetic dynamic graph datasets demonstrate the superiority of our model over state-of-the-art baselines in handling OOD shifts.