DrKGC: Dynamic Subgraph Retrieval-Augmented LLMs for Knowledge Graph Completion across General and Biomedical Domains

2025-05-31

AI Summary

Large language models (LLMs) struggle to effectively perceive and reason over knowledge graph (KG) structures, limiting knowledge graph completion (KGC) performance. To address this, we propose a structure-aware LLM reasoning framework. Our method integrates logical rule-guided, bottom-up dynamic subgraph retrieval to precisely extract task-relevant local KG structures, and a GCN-adapter-driven structural embedding enhancement technique that enables fine-grained topological modeling and interpretable reasoning within LLMs. The framework is lightweight and efficient, unifying rule learning, structural embedding, and retrieval-augmented fine-tuning. Evaluated on four standard benchmarks—including two general-domain and two biomedical datasets—our approach achieves significant improvements over state-of-the-art methods. Case studies in the biomedical domain further demonstrate high accuracy and clinically meaningful interpretability.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Machine Learning: Structured Learning
• Knowledge Representation and Reasoning: Description Logics

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
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Knowledge graph completion (KGC) aims to predict missing triples in knowledge graphs (KGs) by leveraging existing triples and textual information. Recently, generative large language models (LLMs) have been increasingly employed for graph tasks. However, current approaches typically encode graph context in textual form, which fails to fully exploit the potential of LLMs for perceiving and reasoning about graph structures. To address this limitation, we propose DrKGC (Dynamic Subgraph Retrieval-Augmented LLMs for Knowledge Graph Completion). DrKGC employs a flexible lightweight model training strategy to learn structural embeddings and logical rules within the KG. It then leverages a novel bottom-up graph retrieval method to extract a subgraph for each query guided by the learned rules. Finally, a graph convolutional network (GCN) adapter uses the retrieved subgraph to enhance the structural embeddings, which are then integrated into the prompt for effective LLM fine-tuning. Experimental results on two general domain benchmark datasets and two biomedical datasets demonstrate the superior performance of DrKGC. Furthermore, a realistic case study in the biomedical domain highlights its interpretability and practical utility.

Problem

• Predict missing triples in knowledge graphs using LLMs
• Enhance LLM graph perception via dynamic subgraph retrieval
• Improve KGC performance in general and biomedical domains

Innovation

• Dynamic subgraph retrieval for KG completion
• Lightweight model learns structural embeddings
• GCN adapter enhances LLM fine-tuning