Bridging Graph Structure and Knowledge-Guided Editing for Interpretable Temporal Knowledge Graph Reasoning

2026-01-29

AI Summary

This work addresses the limitations of current large language models (LLMs) in temporal knowledge graph reasoning, which often neglect structural relationships, leading to structurally weak, hallucinatory, and temporally inconsistent predictions. To overcome this, the authors propose IGETR, a novel framework that uniquely integrates a structure-aware temporal graph neural network (Temporal GNN) with a knowledge-guided LLM. The approach first employs the Temporal GNN to extract temporally consistent candidate reasoning paths and then leverages the LLM to refine logical and semantic inaccuracies, yielding precise and interpretable predictions. Evaluated on standard benchmarks such as ICEWS, IGETR achieves state-of-the-art performance, improving Hits@1 and Hits@3 by 5.6% and 8.1%, respectively. Ablation studies further confirm the contribution of each component to the overall effectiveness.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Natural Language Processing: Information Extraction
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Temporal knowledge graph reasoning (TKGR) aims to predict future events by inferring missing entities with dynamic knowledge structures. Existing LLM-based reasoning methods prioritize contextual over structural relations, struggling to extract relevant subgraphs from dynamic graphs. This limits structural information understanding, leading to unstructured, hallucination-prone inferences especially with temporal inconsistencies. To address this problem, we propose IGETR (Integration of Graph and Editing-enhanced Temporal Reasoning), a hybrid reasoning framework that combines the structured temporal modeling capabilities of Graph Neural Networks (GNNs) with the contextual understanding of LLMs. IGETR operates through a three-stage pipeline. The first stage aims to ground the reasoning process in the actual data by identifying structurally and temporally coherent candidate paths through a temporal GNN, ensuring that inference starts from reliable graph-based evidence. The second stage introduces LLM-guided path editing to address logical and semantic inconsistencies, leveraging external knowledge to refine and enhance the initial paths. The final stage focuses on integrating the refined reasoning paths to produce predictions that are both accurate and interpretable. Experiments on standard TKG benchmarks show that IGETR achieves state-of-the-art performance, outperforming strong baselines with relative improvements of up to 5.6% on Hits@1 and 8.1% on Hits@3 on the challenging ICEWS datasets. Additionally, we execute ablation studies and additional analyses confirm the effectiveness of each component.

Problem

• Temporal Knowledge Graph Reasoning
• Graph Structure
• LLM-based Reasoning
• Temporal Inconsistency
• Hallucination

Innovation

• Temporal Knowledge Graph Reasoning
• Graph Neural Networks
• Large Language Models
• Path Editing
• Interpretable Reasoning