Research on Comparison and Fusion between Large Language Models and Knowledge Graphs
[Purpose/significance]This paper aims to explore the complementarity in development between two important knowl-edge representation technologies,large language models and knowledge graphs,so as to provide guidance for their integrated application.[Method/process]Through the research of literature data,the paper conducted an in-depth study on the construction methods,knowl-edge representation,and application scenarios of large language models and knowledge graphs.It used the comparative analysis to reveal their similarities and differences in technical features,knowledge representation,and application domains.Finally,it summarized the complementarity of the two and the possibility of integrated application.[Result/conclusion]Enhancing large language models through knowledge graphs will help to improve their pre-training and inference capabilities.Enhancing the construction,completion and question answering ability of knowledge graphs through large language models can achieve more comprehensive,accurate and intelligent knowl-edge service.
large language modelknowledge graphartificial intelligencenatural language processknowledge representation
万方数据
维普
