Research on the Entity Relationship Extraction Method of Large Model Chinese Electronic Medical Record With Low-Moment Features

The entity relationship extraction of Chinese electronic medical records is used to construct the knowledge graph in the medical field, and it plays an important role in serving downstream tasks. However, due to the fuzzy boundary of Chinese electronic medical records, the complex relationship between medical texts, and the high density of entities, the task of extracting the physical relationships of medical texts is not accurate enough. To solve this problem, this paper adopts a large language model ChatGLM2-6B based on (Generative Language Model) architecture and Quantized Low-rank Adaptation of Large Language Models (QLoRA) for Chinese electronic medical record entity relationship extraction. By leveraging the powerful language understanding capabilities of ChatGLM2-6B and the efficient fine-tuning characteristics of QLoRA, the low-moment features are integrated to achieve accurate and efficient Chinese electronic medical record entity relationship extraction, which provides strong support for medical data analysis and clinical decision-making.