This survey synthesizes methodologies for integrating large language models and knowledge graphs, highlighting strengths and challenges in question answering.
Key Points
This survey identifies a structured taxonomy for integrating large language models and knowledge graphs for question answering.
State-of-the-art methods for synthesizing LLMs and KGs show varying strengths and limitations in complex question answering tasks.
Systematic analysis of current methodologies emphasizes key challenges related to reasoning capacity and knowledge updates.
Open challenges in the integration of LLMs and KGs highlight opportunities for further research in question answering.