DPCL-Diff: The Temporal Knowledge Graph Reasoning Based on Graph Node Diffusion Model with Dual-Domain Periodic Contrastive Learning

TL;DR

This paper introduces DPCL-Diff, a novel approach combining graph node diffusion and dual-domain contrastive learning to improve temporal knowledge graph reasoning, especially for sparse and periodic events, achieving superior prediction accuracy.

Contribution

The paper proposes a new diffusion-based generative model and dual-domain contrastive learning framework for TKG reasoning, addressing data sparsity and periodic event challenges.

Findings

Outperforms state-of-the-art TKG models on four datasets

Enhances reasoning for sparse and periodic events

Demonstrates the effectiveness of combining GNDiff and DPCL

Abstract

Temporal knowledge graph (TKG) reasoning that infers future missing facts is an essential and challenging task. Predicting future events typically relies on closely related historical facts, yielding more accurate results for repetitive or periodic events. However, for future events with sparse historical interactions, the effectiveness of this method, which focuses on leveraging high-frequency historical information, diminishes. Recently, the capabilities of diffusion models in image generation have opened new opportunities for TKG reasoning. Therefore, we propose a graph node diffusion model with dual-domain periodic contrastive learning (DPCL-Diff). Graph node diffusion model (GNDiff) introduces noise into sparsely related events to simulate new events, generating high-quality data that better conforms to the actual distribution. This generative mechanism significantly enhances the model's ability to reason about new events. Additionally, the dual-domain periodic contrastive learning (DPCL) maps periodic and non-periodic event entities to Poincar\'e and Euclidean spaces, leveraging their characteristics to distinguish similar periodic events effectively. Experimental results on four public datasets demonstrate that DPCL-Diff significantly outperforms state-of-the-art TKG models in event prediction, demonstrating our approach's effectiveness. This study also investigates the combined effectiveness of GNDiff and DPCL in TKG tasks.