Exploring Large Language Models for Knowledge Graph Completion

TL;DR

This paper presents a novel framework called KG-LLM that leverages large language models for knowledge graph completion, achieving state-of-the-art results by using entity and relation descriptions as prompts.

Contribution

Introducing KG-LLM, a new method that models knowledge graph triples as text sequences and employs LLMs for improved completion performance.

Findings

KG-LLM achieves state-of-the-art results on benchmark tasks.

Fine-tuning smaller models outperforms larger models like GPT-4.

Using entity and relation descriptions as prompts enhances prediction accuracy.

Abstract

Knowledge graphs play a vital role in numerous artificial intelligence tasks, yet they frequently face the issue of incompleteness. In this study, we explore utilizing Large Language Models (LLM) for knowledge graph completion. We consider triples in knowledge graphs as text sequences and introduce an innovative framework called Knowledge Graph LLM (KG-LLM) to model these triples. Our technique employs entity and relation descriptions of a triple as prompts and utilizes the response for predictions. Experiments on various benchmark knowledge graphs demonstrate that our method attains state-of-the-art performance in tasks such as triple classification and relation prediction. We also find that fine-tuning relatively smaller models (e.g., LLaMA-7B, ChatGLM-6B) outperforms recent ChatGPT and GPT-4.