Research and Application of Electronic Data Retrieval in Material Supply Chain Enhanced by Large Language Models and Knowledge Graph
Xing Ge, Yafei Liu, Pei Yang, Xin Han Sun, Junfeng Qiao, Luyao Qu, Jingyi Qiu · 2024
In response to the new goals of building green and modern smart supply chains, the electric power equipment supply chain is experiencing a shift toward digital intelligence and low-carbon, environmentally friendly development [1]. However, traditional search platforms based on relational databases face challenges in handling vast amounts of multimodal electronic data. These platforms often suffer from low search accuracy, limited cross-dimensional correlation analysis capabilities, and inefficiencies, making them inadequate for constructing comprehensive big data platforms that integrate and share information across the entire green, modern, smart supply chain. This paper introduces an innovative electronic data retrieval method designed to meet the data retrieval needs of material supply chains. By integrating large language models with knowledge graph technology, it proposes an electronic data retrieval system that leverages vector database technology for text embedding of multimodal data. Additionally, artificial intelligence is used to enable knowledge retrieval and augmented generation, significantly enhancing data retrieval capabilities within the specialized domain of material supply chains.