The Deep Integration of Knowledge Graphs and Large Language Models: Advancements, Challenges, and Future Directions
Daifeng Li, Fan Xu · 2024
In recent years, large language models have emerged with amazing capabilities, but they also have limitations such as hallucinations and black boxes, while knowledge graphs have accurate knowledge and symbolic reasoning capabilities. Therefore, the integration of knowledge graph and large language model becomes necessary. The paper systematically conducts the recent research pertaining to the integration of knowledge graph with large language mode. By analyzing the methods and locations of their confluence, we propose a unifying framework that aims to facilitate comprehension and inspiration for fellow researchers in related fields. This framework not only consolidates existing knowledge but also enhances the translational potential of research by delineating innovative pathways for practical implementation.