Article Source

• Unifying LLMs and Knowledge Graphs

Unifying LLMs and Knowledge Graphs: A Roadmap and Recent Advances

Abstract

Large language models (LLMs) are making new waves in the field of natural language processing and artificial intelligence. However, LLMs are black-box models, which often fall short of capturing and accessing factual knowledge. In contrast, Knowledge Graphs (KGs) 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 evolve 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 talk, we will present a forward-looking roadmap for the unification of LLMs and KGs. Specifically, we will discuss the recent advances in this direction, including 1) Applying KGs to interpretable the knowledge and reasoning inside LLMs, 2) Adopting LLMs to augment the reasoning on KGs, and 3) Enhancing fiathful LLM reasoning with the assistance of KGs. We will also discuss the opportunities and potential future research directions.

Bio

Linhao Luo is a second-year Ph.D. student of computer science at Monash University. His research interests mainly focus on the areas of artificial intelligence and data mining, especially for the knowledge graph, large language models, graph neural networks, and recommender systems. Previously, He has published several papers in top-tier conferences and journals, e.g., ICLR, SIGIR, ICDE, WSDM, IJCAI, CIKM, and TKDE. He has also worked at Tencent and Alibaba as a research intern.