[论文解读] Unifying Large Language Models and Knowledge Graphs: A Roadmap

Shirui Pan, Linhao Luo+4 more|arXiv (Cornell University)|Jun 14, 2023

Topic Modeling参考文献 257被引用 100

一句话总结

该论文提供一个三框架路线图，用于统一 LLMs 与知识图谱：KG-enhanced LLMs、LLM-augmented KGs，以及 synergized LLMs + KGs，并包含详细的分类与未来方向。

ABSTRACT

Large language models (LLMs), such as ChatGPT and GPT4, are making new waves in the field of natural language processing and artificial intelligence, due to their emergent ability and generalizability. However, LLMs are black-box models, which often fall short of capturing and accessing factual knowledge. In contrast, Knowledge Graphs (KGs), Wikipedia and Huapu for example, are structured knowledge models that explicitly store rich factual knowledge. KGs can enhance LLMs by providing external knowledge for inference and interpretability. Meanwhile, KGs are difficult to construct and evolving by nature, which challenges the existing methods in KGs to generate new facts and represent unseen knowledge. Therefore, it is complementary to unify LLMs and KGs together and simultaneously leverage their advantages. In this article, we present a forward-looking roadmap for the unification of LLMs and KGs. Our roadmap consists of three general frameworks, namely, 1) KG-enhanced LLMs, which incorporate KGs during the pre-training and inference phases of LLMs, or for the purpose of enhancing understanding of the knowledge learned by LLMs; 2) LLM-augmented KGs, that leverage LLMs for different KG tasks such as embedding, completion, construction, graph-to-text generation, and question answering; and 3) Synergized LLMs + KGs, in which LLMs and KGs play equal roles and work in a mutually beneficial way to enhance both LLMs and KGs for bidirectional reasoning driven by both data and knowledge. We review and summarize existing efforts within these three frameworks in our roadmap and pinpoint their future research directions.

研究动机与目标

• 提出一个前瞻性的路线图，以统一 LLMs 与 KGs 并发挥它们互补的优势。
• 提供在每个集成框架内的细粒度分类与现有工作评述。
• 总结 LLMs 与 KGs 的进展，包括多模态知识图谱和最先进的模型。
• 突出挑战并勾勒未来研究的有前景方向。

提出的方法

• 定义用于统一 LLMs 与 KGs 的三个通用框架：KG-enhanced LLMs、LLM-augmented KGs、以及 Synergized LLMs + KGs。
• 在每个框架内发展细粒度分类（pre-training、inference、interpretability；embedding、completion、construction、KG-to-text、QA；knowledge representation and reasoning）。
• 评估跨知识整合技术、提示、检索与指令微调的现有方法与分类体系。
• 综合挑战与未来研究方向，以指导该领域后续工作。

实验结果

研究问题

• RQ1 如何将知识图谱与大语言模型整合，以克服幻觉与可解释性问题？
• RQ2 在 pre-training、inference 与 instruction-tuning 期间，哪些整合策略最能利用 KG 结构？
• RQ3 LLMs 如何增强 KG 任务，如 embedding、completion、construction，以及 KG-to-text 生成？
• RQ4 在双向推理中协同 LLMs 与 KGs 的关键挑战与未来方向是什么？

主要发现

• 提供一个结构化的路线图，包含三个框架以统一 LLMs 与 KGs。
• 提供每个框架内研究的细粒度分类法，并总结代表性方法。
• 讨论最先进的 LLMs 与不断发展的 KGs，包括多模态知识图谱。
• 识别知识更新、可解释性和零-shot 迁移等挑战，并提出未来工作的方向。