T3DM: Test-Time Training-Guided Distribution Shift Modelling for Temporal Knowledge Graph Reasoning

TL;DR

This paper introduces T3DM, a novel approach for temporal knowledge graph reasoning that models distribution shifts during test time and improves negative sampling quality, leading to more robust and accurate results.

Contribution

The paper proposes T3DM, a test-time training method that models distribution shifts and enhances negative sampling for better temporal knowledge graph reasoning.

Findings

T3DM outperforms state-of-the-art baselines in most cases.

It effectively models distribution shifts between training and test data.

The negative sampling strategy improves the quality of negative quadruples.

Abstract

Temporal Knowledge Graph (TKG) is an efficient method for describing the dynamic development of facts along a timeline. Most research on TKG reasoning (TKGR) focuses on modelling the repetition of global facts and designing patterns of local historical facts. However, they face two significant challenges: inadequate modeling of the event distribution shift between training and test samples, and reliance on random entity substitution for generating negative samples, which often results in low-quality sampling. To this end, we propose a novel distributional feature modeling approach for training TKGR models, Test-Time Training-guided Distribution shift Modelling (T3DM), to adjust the model based on distribution shift and ensure the global consistency of model reasoning. In addition, we design a negative-sampling strategy to generate higher-quality negative quadruples based on adversarial training. Extensive experiments show that T3DM provides better and more robust results than the state-of-the-art baselines in most cases.